about the book… The management of tumors in and adjacent to the skullbase is challenging given the complex and critically important anatomy of the region and the wide diversity of tumor pathologies that may be encountered. To help navigate the complexities of contemporary multidisciplinary management of these patients, Drs. Hanna and DeMonte bring you Comprehensive Management of Skull Base Tumors, a comprehensive guide filled with updated information from authorities around the world. Comprehensive Management of Skull Base Tumors is divided into three sections consisting of: • general principles • site specific surgery • tumor specific management Filled with scientific tables and lavishly illustrated, this text is written with an emphasis on surgery, radiation and chemotherapy, and will appeal to all neurosurgeons, otolaryngologists, plastic surgeons, maxillofacial surgeons, ophthalmologists, medical and radiation oncologists, and radiologists. about the editors... EHAB Y. HANNA is Professor of Head and Neck Surgery and Neurosurgery, The University of Texas M.D. Anderson Cancer Center, and Adjunct Professor of Otolaryngology, Baylor College of Medicine, Houston, Texas, USA. Dr. Hanna is also Vice Chair for Clinical Affairs and Medical Director of the Head and Neck Center at M.D. Anderson Cancer Center. He obtained his M.D. from Ain Shams University School of Medicine, Cairo, Egypt. Dr. Hanna currently serves on the editorial boards of several publications; he is Editor-in-Chief of Head and Neck and Editor of the Head and Neck Cancers Section of Current Oncology Reports. He is an involved member of the American Head and Neck Society, the North American Skull Base Society, the American Academy of Otolaryngology-Head and Neck Surgery, the American College of Surgeons, and the National Cancer Institute. Dr. Hanna has been nationally recognized as one of “America’s Top Doctors” and “America’s Top Doctors for Cancer”. He has been a frequent invited guest speaker at national and international conferences and lectures and has contributed to numerous peer-reviewed publications in the fields of head and neck cancer and skull base tumors. FRANCO DEMONTE is Professor of Neurosurgery and Head and Neck Surgery, Deputy Chairman of the Department of Neurosurgery, and Medical Director of the Brain and Spine Center, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, USA. Dr. DeMonte is also an adjunct professor of neurosurgery at Baylor College of Medicine, Houston, Texas, USA. He received his M.D. from the University of Western Ontario, London, Canada. He is an active member of several organizations, including the Society of Neurological Surgeons, the American Association of Neurological Surgeons, and the Canadian Neurosurgical Society. He has served as president of the North American Skull Base Society and of the Houston Neurological Society. His clinical and educational activities have been recognized through his national and international presentations as well as by several teaching awards and inclusion in “Best Doctors in America” and “America’s Top Doctors”. Printed in the United States of America
Comprehensive Management of Skull Base Tumors
Otolaryngology/Neurosurgery
Comprehensive Management of Skull Base Tumors Edited by
Ehab Y. Hanna & Franco DeMonte
Hanna DK054X
•
DeMonte
Comprehensive Management of Skull Base Tumors
Comprehensive Management of Skull Base Tumors Edited by
Ehab Y. Hanna
The University of Texas M.D. Anderson Cancer Center Houston, Texas, USA
Franco DeMonte
The University of Texas M.D. Anderson Cancer Center Houston, Texas, USA
Informa Healthcare USA, Inc. 52 Vanderbilt Avenue New York, NY 10017 C
2009 by Informa Healthcare USA, Inc. Informa Healthcare is an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-10: 0-8493-4054-3 (Hardcover) International Standard Book Number-13: 978-0-8493-4054-3 (Hardcover) This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequence of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data Comprehensive management of skull base tumors / edited by Ehab Y. Hanna, Franco DeMonte. p. ; cm. Includes bibliographical references and index. ISBN-13: 978-0-8493-4054-3 (hardcover : alk. paper) ISBN-10: 0-8493-4054-3 (hardcover : alk. paper) 1. Skull base–Cancer. 2. Skull base–Tumors–Surgery. I. Hanna, Ehab Y. II. DeMonte, Franco. [DNLM: 1. Skull Base Neoplasms. WE 707 C7374 2008] RD662.5.C66 2008 616.99 481–dc22 2008036404
For Corporate Sales and Reprint Permissions call 212-520-2700 or write to: Sales Department, 52 Vanderbilt Avenue, 16th floor, New York, NY 10017. Visit the Informa Web site at www.informa.com and the Informa Healthcare Web site at www.informahealthcare.com
This book is dedicated to My wife, Sylvie, for her grace, gentle spirit, and beauty; Our daughters, Gabrielle Grace “Gigi” Hanna and Camille Lauren Hanna, for the joy and blessing they bring to our lives; My parents, who encouraged me to follow my dreams; My Fellows, residents, and students, who continue to teach me; And my patients, Whose endurance, resilience, and faith continue to amaze me. Ehab Y. Hanna, MD
My parents, Dolinda and Giacomo, for their love, their sacrifices, and their dedication to my education. To my children, Evan and Madeline who fill me with pride. And, with limitless love, to Paula, my wife and best friend. Franco DeMonte, MD
Preface
Section two covers site-specific information regarding the various anatomic regions of the cranial base, including surgical anatomy, regional pathology, differential diagnosis, clinical assessment, diagnostic imaging, and surgical approaches. This regional classification includes the anterior cranial fossa, sinonasal region, nasopharynx, clivus, infratemporal fossa, parapharyngeal space, temporal bone, sella turcica, middle cranial fossa, petrous apex, cerebellopontine angle, jugular foramen, and craniovertebral junction. Section three covers comprehensive multidisciplinary discussion of tumor-specific topics such as tumor incidence and epidemiology, pathology, staging, treatment, outcome, and prognosis. This section covers the following tumors of the cranial base: squamous and nonsquamous cell carcinoma, olfactory neuroblastoma, melanoma, sarcomas, angiofibromas and other vascular tumors, chordomas and chondrosarcomas, meningiomas, schwannomas, paragangliomas, pituitary adenomas, craniopharyngiomas, epidermoid and dermoid cysts, fibro-osseous lesions, and metastatic tumors. This organizational schema is intended to provide a simple yet comprehensive way for readers to find the information they need. For example, the reader who wants to know about the latest advances in radiation therapy of skull base tumors is referred to the first section, a reader who has a patient with a tumor of the petrous apex is referred to the second section, and another who wants to know all the available treatment options and prognostic factors for esthesioneuroblastoma is referred to the third section. With such an organizational structure, some redundancy is unavoidable, but not, as you will see, detrimental. As with any textbook, some omissions are inevitable and we hope our readership will forgive any shortcomings of this work. We believe that the greatest value of this book is the incredible expertise of the contributing authors. They truly represent the world’s experts on their specific topics. We are honored by their contribution and humbled by their graciousness to join us in this work.
The management of patients with tumors of the skull base has evolved significantly in the last two decades. Major advances have been achieved in the surgical management of these patients, particularly in the areas of tumor resection and surgical reconstruction. These advances can be mainly attributed to the collaborative efforts of dedicated teams representing various surgical disciplines including neurosurgery, head and neck surgery, neuro-otology, oromaxillofacial surgery, ophthalmology, and plastic and reconstructive surgery. Meanwhile significant advances in radiation delivery using various methods of conformal therapy, including three-dimensional computerized tomography (3D-CT), intensity-modulated radiation therapy (IMRT), proton beam therapy, and stereotactic radiation, as well as advances in chemotherapy and targeted biologic therapy have added significantly to the menu of treatment options for patients with tumors of the skull base. Although there are many excellent references describing the surgical management of patients with tumors of the cranial base, this textbook is intended to be a comprehensive guide to help navigate the complexity of contemporary multidisciplinary management of these patients. In addition, we hope that this reference will also provide the reader with deeper understanding of the unique biologic behavior and the underlying molecular and genetic aberrations of the various tumor types originating from or involving the cranial base, and the potential for these molecular derangements to be putative targets for future development of more effective biologic therapy. To address these goals, we have organized the book in three sections: general principles, site-specific chapters, and tumor-specific chapters. Section one covers general topics pertinent to all patients with neoplasms of the skull base, regardless of specific location or tumor type. These topics include anatomy, pathology, genetics, clinical evaluation, diagnostic imaging, anesthesia, minimally invasive surgery, surgical reconstruction, prosthetic rehabilitation, radiation and radiobiology, chemotherapy, evaluation and rehabilitation of speech and swallowing, functional outcomes and quality of life issues, neurocognitive assessment, and cerebrovascular management.
Ehab Y. Hanna Franco DeMonte
v
Contents
Preface . . . . v Contributors . . . . ix
12. Rehabilitation of Speech and Swallowing of Patients with Tumors of the Skull Base 181 Gail L. Davie, Denise A. Barringer, and Jan S. Lewin
Section 1: General Principles
13. Quality of Life of Patients with Skull Base Tumors 189 Ziv Gil and Dan M. Fliss
1. Anatomy of the Cranial Base 3 Carolina Martins and Albert L. Rhoton, Jr.
14. Neurocognitive Assessment of Patients with Tumors of the Skull Base 201 Mariana Witgert and Tracy Veramonti
2. Pathology of Tumor and Tumor-like Lesions of the Skull Base 43 Michelle D. Williams and Adel K. El-Naggar
15. Cerebrovascular Management in Skull Base Tumors 207 Sabareesh Kumar Natarajan, Basavaraj Ghodke, and Laligam N. Sekhar
3. Genetic Abnormalities of Skull Base Tumors 63 Ziv Gil and Dan M. Fliss 4. Imaging of Skull Base Neoplasms 81 Lawrence E. Ginsberg
Section 2: Site-Specific Considerations
5. Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base: Clinical Examination, Auditory Testing, Vestibular Testing, and Equilibrium 95 Paul W. Gidley
16. Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base 227 Ehab Y. Hanna, Michael Kupferman, and Franco DeMonte
6. Anesthesia and Intraoperative Monitoring of Patients with Tumors of the Skull Base 119 Walter S. Jellish and Steven B. Edelstein
17. Tumors of the Nasopharynx 267 William Ignace Wei and Paul K. Y. Lam 18. Clival Tumors 277 Franco DeMonte, Mark J. Dannenbaum, and Ehab Y. Hanna
7. Minimally Invasive Techniques: Endonasal Endoscopic Skull Base Surgery 131 Allan D. Vescan, Ricardo L. Carrau, Carl H. Snyderman, Amin B. Kassam, Arlan Mintz, and Paul Gardner
19. Tumors of the Anterior Skull Base 293 Vijayakumar Javalkar, Bharat Guthikonda, Prasad Vannemreddy, and Anil Nanda
8. Reconstruction of Skull Base Defects 139 Peter C. Neligan, Christine B. Novak, and Patrick J. Gullane
20. Infratemporal/Middle Fossa Tumors 305 Paul J. Donald
9. Prosthetic Rehabilitation of Patients Undergoing Skull Base Surgery 149 Theresa M. Hofstede, Rhonda F. Jacob, Pattii C. Montgomery, Peggy J. Wesley, Jack W. Martin, and Mark S. Chambers
21. Tumors of the Parapharyngeal Space 331 Eric J. Moore and Kerry D. Olsen 22. Tumors of the Temporal Bone 345 Sam J. Marzo and John P. Leonetti
10. Radiobiology and Radiation Therapy of Skull Base Tumors 159 Simon S. Lo, John H. Suh, and Eric L. Chang
23. Evaluation and Management of Sellar Tumors 355 Jay Jagannathan, Edward R. Laws, and John A. Jane, Jr.
11. Chemotherapy for Skull Base Tumors 175 Bilal Ahmed, Ehab Y. Hanna, and Merrill S. Kies
24. Tumors of the Middle Cranial Fossa 367 Ali A. Baaj, Siviero Agazzi, and Harry R. van Loveren
vii
viii
Contents
25. Tumors of the Petrous Apex 375 Ricardo Ramina, Maur´ıcio Coelho Neto, Yvens Barbosa Fernandes, Erasmo Barros da Silva, Jr., and Kristofer Luiz Fingerle Ramina 26. Tumors of the Cerebellopontine Angle 389 Bryan C. Oh, Daniel J. Hoh, and Steven L. Giannotta 27. Tumors of the Jugular Foramen 403 Samer Ayoubi, Badih Adada, and Ossama Al-Mefty 28. Tumors of the Craniovertebral Junction 417 Douglas Fox, Scott Wait, Steve Chang, G. Vini Khurana, Curtis A. Dickman, Volker K. H. Sonntag, and Robert F. Spetzler
Section 3: Tumor-Specific Considerations 29. Squamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses 429 Patrick Sheahan, Snehal G. Patel, and Jatin P. Shah 30. Nonsquamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses 445 ` Carlo L. Solero, Stefano Riccio, and Giulio Cantu, Sarah Colombo 31. Esthesioneuroblastoma 453 Valerie J. Lund and David J. Howard 32. Melanoma of the Nasal Cavity and Paranasal Sinuses 459 Ziv Gil, Mark H. Bilsky, and Dennis H. Kraus 33. Sarcomas of the Skull Base 473 Katherine A. Thornton and Robert S. Benjamin
34. Angiofibromas and Vascular Tumors of the Skull Base 481 Andrew G. Sikora and Randal S. Weber 35. Chordoma and Chondrosarcoma of the Skull Base 495 Gordon T. Sakamoto and Griffith R. Harsh 36. Meningioma 503 Ashwin Viswanathan and Franco DeMonte 37. Schwannomas of the Skull Base 513 Daniel W. Nuss and Emily Lifsey Burke 38. Paragangliomas of the Head and Neck 539 David P. Goldstein, Mark G. Shrime, Bernard Cummings, and Patrick J. Gullane 39. Pituitary Adenomas 557 Mark Hornyak and William T. Couldwell 40. Craniopharyngioma: Neurosurgical Management 573 Douglas James Cook and James T. Rutka 41. Epidermoids, Dermoids, and Other Cysts of the Skull Base 583 Samuel P. Gubbels and Bruce J. Gantz 42. Fibro-Osseous Lesions of the Skull Base 597 Ian T. Jackson 43. Metastatic Skull Base Tumors 615 Krishna Satyan and Sujit S. Prabhu Index . . . . 619
Contributors
Maur´ıcio Coelho Neto Department of Neurosurgery, Neurological Institute of Curitiba, Curitiba, Brazil
Badih Adada Department of Neurosurgery, The University of Arkansas for Medical Sciences, Little Rock, Arkansas, U.S.A.
Sarah Colombo Cranio-Maxillo-Facial Department, National Cancer Institute, Milano, Italy
Siviero Agazzi Department of Neurosurgery, University of South Florida, Tampa, Florida, U.S.A.
Douglas James Cook Division of Neurosurgery, University of Toronto, Toronto, Ontario, Canada
Bilal Ahmed Departments of Thoracic/Head and Neck Medical Oncology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
William T. Couldwell Department of Neurosurgery, University of Utah, Salt Lake City, Utah, U.S.A.
Ossama Al-Mefty Department of Neurosurgery, University of Arkansas for Medical Sciences, Little Rock, Arkansas, U.S.A.
Bernard Cummings Department of Radiation Oncology, Princess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada
Samer Ayoubi Department of Neurosurgery, Abbassi Medical Center, Damascus, Syria
Mark J. Dannenbaum Department of Neurosurgey, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Ali A. Baaj Department of Neurosurgery, University of South Florida, Tampa, Florida, U.S.A.
Gail L. Davie Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Denise A. Barringer Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Franco DeMonte Department of Neurosurgey, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Erasmo Barros da Silva, Jr. Department of Neurosurgery, Neurological Institute of Curitiba, Curitiba, Brazil, Postgraduate Course in Surgery, Pontifical Catholic University of Parana, Curitiba, Brazil
Curtis A. Dickman Division of Neurological Surgery, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, Phoenix, Arizona, U.S.A.
Robert S. Benjamin Department of Sarcoma Medical Oncology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Paul J. Donald Department of Otolaryngology-Head and Neck Surgery, University of California-Davis, Sacramento, California, U.S.A.
Mark H. Bilsky Department of Neurosurgery, Memorial Sloan-Kettering Cancer Center, Cornell University Medical College, New York, New York, U.S.A.
Steven B. Edelstein Loyola University Medical Center and Loyola University Stritch School of Medicine, Department of Anesthesiology, Maywood, Illinois, U.S.A.
Emily Lifsey Burke Department of Otolaryngology-Head & Neck Surgery, Louisiana State University Health Sciences Center, New Orleans and Baton Rouge, Louisiana, U.S.A.
Adel K. El-Naggar Department of Pathology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Giulio Cantu´ Cranio-Maxillo-Facial Department, National Cancer Institute, Milano, Italy
Yvens Barbosa Fernandes Department of Neurosurgery, Neurological Institute of Curitiba, Department of Neurosurgery, State University of Campinas (UNICAMP), Campinas, Brazil
Ricardo L. Carrau Department of Otolaryngology and Head & Neck Surgery, Department of Neurological Surgery, and Minimally Invasive Endoneurosurgery Center, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, U.S.A.
Dan M. Fliss Department of Otolaryngology Head and Neck Surgery, Tel-Aviv Sourasky Medical Center, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
Mark S. Chambers Section of Oncologic Dentistry and Prosthodontics, Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Douglas Fox
Private practice, Austin, Texas, U.S.A.
Bruce J. Gantz Department of Otolaryngology/Head and Neck Surgery, University of Iowa Hospitals and Clinics, Iowa City, Iowa, U.S.A.
Eric L. Chang Department of Radiation Oncology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Paul Gardner Department of Neurological Surgery and Minimally Invasive Endoneurosurgery Center, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, U.S.A.
Steve Chang Division of Neurological Surgery, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, Phoenix, Arizona, U.S.A. ix
x
Contributors
Paul W. Gidley Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Vijayakumar Javalkar Department of Neurosurgery, Louisiana State University Health Sciences Center — Shreveport, Louisiana, U.S.A.
Basavaraj Ghodke Departments of Neurological Surgery, and Neuroradiology, University of Washington, Seattle, Washington, U.S.A.
Walter S. Jellish Department of Anesthesiology, Loyola University Hospital Medical Center, Maywood, Illinois, U.S.A.
Steven L. Giannotta Department of Neurological Surgery, Keck School of Medicine, University of Southern California, Los Angeles, California, U.S.A.
Amin B. Kassam Department of Neurological Surgery, Department of Otolaryngology and Head & Neck Surgery, and Minimally Invasive Endoneurosurgery Center, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, U.S.A.
Ziv Gil Department of Otolaryngology Head and Neck Surgery, Tel-Aviv Sourasky Medical Center, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel Lawrence E. Ginsberg Department of Radiology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A. David P. Goldstein Department of Surgical Oncology, Princess Margaret Hospital, Toronto, Ontario, Department of Otolaryngology– Head and Neck Surgery, University of Toronto, Toronto, Ontario, Canada Samuel P. Gubbels Department of Surgery, Division of Otolaryngology/Head and Neck Surgery, University of Wisconsin, Madison, Wisconsin, U.S.A. Patrick J. Gullane Department of Otolaryngology– Head and Neck Surgery, University of Toronto, Department of Surgical Oncology, Princess Margaret Hospital, Toronto, Ontario, Canada Bharat Guthikonda Department of Neurosurgery, Louisiana State University, Health Sciences Center—Shreveport, Louisiana, U.S.A.
G. Vini Khurana The Canberra Hospital, Department of Neurosurgery, Canberra Medical School, Australian National University, Canberra, Australian Capital Territory, Australia Merrill S. Kies Departments of Thoracic/Head and Neck Medical Oncology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A. Dennis H. Kraus Head and Neck Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, Cornell University Medical College, New York, New York, U.S.A. Michael Kupferman Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A. Paul K. Y. Lam Division of Otorhinolaryngology, Head and Neck Surgery, University of Hong Kong Medical Centre, Queen Mary Hospital, Hong Kong SAR, P.R. China Edward R. Laws Department of Neurosurgery, Harvard University, Boston, Massachusetts, U.S.A.
Ehab Y. Hanna Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
John P. Leonetti Department of Otolaryngology, Head and Neck Surgery, Loyola University Health System, Maywood, Illinois, U.S.A.
Griffith R. Harsh Department of Neurosurgery, Stanford University, Palo Alto, California, U.S.A.
Jan S. Lewin Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Theresa M. Hofstede Section of Oncologic Dentistry and Prosthodontics, Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A. Daniel J. Hoh Department of Neurological Surgery, Keck School of Medicine, University of Southern California, Los Angeles, California, U.S.A. Mark Hornyak Department of Neurosurgery, University of Utah, Salt Lake City, Utah, U.S.A.
Simon S. Lo Department of Radiation Medicine, Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Ohio State University Medical Center, Columbus, Ohio, U.S.A. Valerie J. Lund Royal National Throat Nose and Ear Hospital, Ear Institute, University College London, London, U.K.
David J. Howard Royal National Throat Nose and Ear Hospital, Ear Institute, University College London, London, U.K.
Jack W. Martin Section of Oncologic Dentistry and Prosthodontics, Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Ian T. Jackson Craniofacial Institute, Southfield, Michigan, U.S.A.
Carolina Martins Medical School of Pernambuco IMIP & ´ Hospital Getulio Vargas, Recife, Brazil
Rhonda F. Jacob Section of Oncologic Dentistry and Prosthodontics, Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Sam J. Marzo Department of Otolaryngology, Head and Neck Surgery, Loyola University Health System, Maywood, Illinois, U.S.A.
Jay Jagannathan Department of Neurosurgery, University of Virginia Health Sciences Center, University of Virginia, Charlottesville, Virginia, U.S.A. John A. Jane, Jr. Department of Neurosurgery, University of Virginia Health Sciences Center, University of Virginia, Charlottesville, Virginia, U.S.A.
Arlan Mintz Department of Neurological Surgery and Minimally Invasive Endoneurosurgery Center, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, U.S.A. Pattii C. Montgomery Section of Oncologic Dentistry and Prosthodontics, Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Contributors
Eric J. Moore Department of Otolaryngology, The Mayo Graduate College of Medicine, The Mayo Clinic, Rochester, Minnesota, U.S.A. Anil Nanda Department of Neurosurgery, Louisiana State University Health Sciences Center—Shreveport, Louisiana, U.S.A. Sabareesh Kumar Natarajan Department of Neurological Surgery, University of Washington, Seattle, Washington, U.S.A. Peter C. Neligan Division of Plastic Surgery, University of Washington, Seattle, Washington, U.S.A. Christine B. Novak Wharton Head and Neck Center, University Health Network, University of Toronto, Toronto, Ontario, Canada Daniel W. Nuss Department of Otolaryngology-Head & Neck Surgery, and Department of Neurological Surgery, Louisiana State University Health Sciences Center, New Orleans and Baton Rouge, Louisiana, U.S.A. Bryan C. Oh Department of Neurological Surgery, Keck School of Medicine, University of Southern California, Los Angeles, California, U.S.A. Kerry D. Olsen Department of Otolaryngology, The Mayo Graduate College of Medicine, The Mayo Clinic, Rochester, Minnesota, U.S.A. Snehal G. Patel Head and Neck Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York, U.S.A. Sujit S. Prabhu Department of Neurosurgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A. Kristofer Luiz Fingerle Ramina Department of ¨ ¨ Neurosurgery, University of Tubingen, Tubingen, Germany Ricardo Ramina Department of Neurosurgery, Neurological Institute of Curitiba, Curitiba, Brazil, Postgraduate Course in Surgery, Pontifical Catholic University of Parana, Curitiba, Brazil Albert L. Rhoton, Jr. Department of Neurosurgery, University of Florida, Gainesville, Florida, U.S.A. Stefano Riccio Cranio-Maxillo-Facial Department, National Cancer Institute, Milano, Italy James T. Rutka Division of Neurosurgery, The Hospital for Sick Children, The University of Toronto, Toronto, Ontario, Canada Gordon T. Sakamoto Stanford University, Palo Alto, California, U.S.A. Krishna Satyan Department of Neurosurgery, The University of Texas M.D. Anderson Cancer Center, Houston Texas, U.S.A.
xi
Mark G. Shrime Department of Surgical Oncology, Princess Margaret Hospital, Toronto, Ontario, Department of Otolaryngology– Head and Neck Surgery, University of Toronto, Toronto, Ontario, Canada Andrew G. Sikora Department of Otolaryngology, Mount Sinai School of Medicine, New York, New York, U.S.A. Carl H. Snyderman Department of Otolaryngology and Head & Neck Surgery, Department of Neurological Surgery, and Minimally Invasive Endoneurosurgery Center, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, U.S.A. Carlo L. Solero Second Neurosurgical Unit, Istituto Nazionale Neurologico “C. Besta”, Milano, Italy Volker K. H. Sonntag Division of Neurological Surgery, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, Phoenix, Arizona, U.S.A. Robert F. Spetzler Division of Neurological Surgery, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, Phoenix, Arizona, U.S.A. John H. Suh Department of Radiation Oncology, The Cleveland Clinic Foundation, Cleveland, Ohio, U.S.A. Katherine A. Thornton Johns Hopkins University School of Medicine, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, U.S.A. Harry R. van Loveren Department of Neurosurgery, University of South Florida, Tampa, Florida, U.S.A. Prasad Vannemreddy Department of Neurosurgery, Louisiana State University Health Sciences Center — Shreveport, Louisiana, U.S.A. Tracy Veramonti Department of Neuro-Oncology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A. Ashwin Viswanathan Department of Neurosurgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A. Allan D. Vescan Department of Otolaryngology and Head & Neck Surgery, and Minimally Invasive Endoneurosurgery Center, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, U.S.A. Scott Wait Division of Neurological Surgery, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, Phoenix, Arizona, U.S.A. Randal S. Weber Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A. William Ignace Wei Division of Otorhinolaryngology, Head and Neck Surgery, University of Hong Kong Medical Centre, Queen Mary Hospital, Hong Kong SAR, P.R. China
Laligam N. Sekhar Department of Neurological Surgery, University of Washington, Seattle, Washington, U.S.A.
Peggy J. Wesley Section of Oncologic Dentistry and Prosthodontics, Department of Head and Neck Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Jatin P. Shah Head and Neck Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York, U.S.A.
Michelle D. Williams Department of Pathology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Patrick Sheahan Head and Neck Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York, U.S.A.
Mariana Witgert Department of Neuro-Oncology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A.
Section 1 General Principles
1 Anatomy of the Cranial Base Carolina Martins and Albert L. Rhoton, Jr.
OVERVIEW
galli and the cribriform plate of the ethmoid bone anteriorly and the planum of the sphenoid body posteriorly. The lateral part, which covers the orbit and the optic canal, is formed by the frontal bone and the lesser wing of the sphenoid bone, which blends medially into the anterior clinoid process (Figs. 3 and 4). The foramen caecum in the midline serves as the site of passage of an emissary vein and the cribriform plate is pierced by the filaments of the olfactory nerve. The optic canal transmits the optic nerve and the ophthalmic artery. The anterior cranial base faces the frontal lobes with the gyri recti medially and the orbital gyri laterally, along with the branches of the anterior cerebral arteries medially and middle cerebral arteries laterally.
No part of the cranial base is immune to surgical pathology or to its use as a pathway to access lesions in the intra- or extracranial spaces. Tumors and multiple other lesions can involve any of the intracranial fossae, and can appear in the paranasal sinuses, nasal cavity, infratemporal and pterygopalatine fossae, orbit, and in the retropharyngeal and craniocervical regions (Fig. 1). Managing these lesions requires an extensive knowledge of the anatomy of the cranial base and its intra- and extracranial relationships. This chapter provides a concise review of the cranial base. The skull is divided into the cranium and the facial skeleton. The cranium is divided into the calvarium and the cranial base. The cranial base has an endocranial surface, which faces the brain, and an exocranial surface, which faces the nasal cavity and sinuses, orbits, pharynx, infratemporal and pterygopalatine fossae, and the parapharyngeal and infrapetrosal spaces (Fig. 2). Both surfaces are connected by canals, foramina, and fissures through which numerous neural and vascular structures pass. Both the endocranial and exocranial cranial base surfaces are divided into anterior, middle, and posterior parts, each of which has a central and paired lateral portions. On the intracranial side, the three parts correspond to the anterior, middle, and posterior cranial fossae (Figs. 2 and 4) (1,2). On the endocranial side, the border between the anterior and middle cranial bases is the sphenoid ridge joined medially by the chiasmatic sulcus, and the border between the middle and posterior cranial bases is formed by the petrous ridges joined by the dorsum sellae and posterior clinoid processes. On the exocranium side, the anterior and middle cranial bases are divided at the level of a transverse line extending through the pterygomaxillary fissures and the pterygopalatine fossae at the upper level and the posterior edge of the alveolar process of the maxilla at a lower level. Medially, this corresponds to the anterior part of the attachment of the vomer to the sphenoid bone. The middle and posterior cranial bases are separated by a transverse line crossing at or near the posterior border of the vomer– sphenoid junction, the foramen lacerum, carotid canal, jugular foramen, styloid process, and the mastoid tip. The osseous structures, their foramina and fissures, canals and their muscular, and neural and vascular relationships are described in this chapter.
Exocranial Surface On the exocranial side, the anterior cranial base is divided in a medial part related to the ethmoid and the sphenoid sinuses with the nasal cavity below and a lateral part that corresponds to the orbit and maxilla (Figs. 2, 6, 7, and 8) (1). The ethmoid bone forms the anterior and middle third and the sphenoid body forms the posterior third of the medial part. The ethmoid is formed by the cribriform plate with the olfactory fila traveling through it, the perpendicular plate, which joins the vomer in forming the nasal septum, and two lateral plates located in the medial walls of the orbits. The lateral plates separate the lateral wall of the nasal cavity and the orbit. The superior turbinate, an appendage of the ethmoid bone, projects into the superior part of the nasal cavity. The body of the sphenoid bone harbors the sphenoid sinus just below the planum sphenoidale, with the anterior orifices located above the superior turbinate. The orbital roof is formed by the lesser sphenoid wing and by the orbital plate of the frontal bone; the lateral wall is formed by the greater sphenoid wing and the zygomatic bone; the inferior wall is formed by the zygomatic, maxillary, and palatine bones; and the medial wall is formed by maxillary, lacrimal, and ethmoid bones (3,4). The main foramina of the region are the anterior and posterior ethmoidal foramen located in the superomedial orbital wall, transmitting the anterior and posterior ethmoidal nerves and arteries; the supraorbital and supratrochlear notches or foramina, transmitting the arteries and nerves of the same name; and the optic canal, through which the optic nerve and ophthalmic artery pass (Figs. 4, 5, and 7). The superior orbital fissure is located between the lesser and greater wing of the sphenoid bone on the lateral side of the optic canal. The inferior orbital fissure, located between the greater sphenoid wing behind and the maxillary and palatine bones anteriorly, is closed by fibrous tissue and orbital muscle. Covered with periorbita and filled with a great amount of fat, the orbit is divided into an anterior space where the globe lies and a posterior space that shelters the nerves, vessels, and muscles behind the globe (5). The
ANTERIOR CRANIAL BASE Endocranial Surface The anterior endocranial surface, formed by the ethmoid, sphenoid, and frontal bones, is divided into medial and lateral portions (Figs. 3–5). The medial part, covering the upper nasal cavity and the sphenoid sinus, is formed by the crista 3
4
Martins and Rhoton
A
Car. A.
B
Orbit roof Sphen. sinus Eth. sinus
V3 V2
Pterygopal. fossa Infratemp. fossa
Car. A.
V2 V3
Sphen. sinus Nasal septum
Nasal cavity
Maxilla
Pterygopal. fossa Nasal cavity
Eth. sinus Maxilla
Orbit roof
C
D
Pit. stalk
Car. A.
Car. A.
Sella Orb. apex
Max. sinus Eust. tube Vomer
Sphen. sinus
V3 Front. N. V2 Max. A. CN II Eth. air cell Pterygopal. fossa Nasopharynx Max. sinus
Septal cart.
Infraorb. N. Max. sinus Nasolac. duct
Sphen. sinus
V3 Pterygopal. gang.
Pteryg. M. Max. A. Br. Vomer Nasal cavity
Maxilla Alv. part
Figure 1
Chapter 1: Anatomy of the Cranial Base
annular tendon of Zinn, a fibrous ring that surrounds the central part of the superior orbital fissure and the optic canal, gives attachment to the superior, medial, inferior, and lateral rectus muscles (Fig. 4). The superior oblique attaches above the annular tendon and the inferior oblique arises from the inferomedial orbital wall just behind the rim. The oculomotor foramen, located inside the annular tendon and through which the oculomotor nerve passes, is located between the upper and lower attachment of the lateral rectus muscle. Just before passing through the superior orbital fissure and the oculomotor foramen in the annular tendon, the oculomotor nerve divides into an upper division supplying the superior rectus and levator muscles and a lower division to the medial and inferior rectus and inferior oblique muscles. The oculomotor nerve gives rise to the parasympathetic motor root to the ciliary ganglion which lies lateral to the optic nerve. The abducens nerve passes through the oculomotor foramen and enters the medial surface of the lateral rectus muscle. The ophthalmic nerve divides just behind the annular tendon into lacrimal and frontal nerves which pass outside the annular tendon, and into the nasociliary nerve which passes through the annular tendon. The ophthalmic nerve gives rise to the long ciliary nerves and the sensory root to the ciliary ganglion; the former conveys the sympathetic pupillomotor fibers and the latter conveys corneal sensation. The trochlear nerve passes above and outside the superomedial edge of the annular tendon. The optic nerve passes superior and medial from the globe to reach the optic canal and divides the retro-orbital space in medial and lateral parts. The main arterial supply to the orbit is by the ophthalmic artery and its branches. This artery courses below the optic nerve in the optic canal, crosses to the lateral side of the nerve at the orbital apex, and then courses from lateral to medial above the optic nerve. The main branches are the central retinal artery and the lacrimal, ciliary, ethmoidal, supraorbital, and dorsal nasal arteries, plus numerous muscular branches. The main venous drainage of the orbit is through the superior and inferior ophthalmic veins that exit the orbit by passing outside the annular tendon and through the superior orbital fissure. The lacrimal gland, located in the superolateral part of the orbit, receives its sensory innervation from the lacrimal nerve, and its parasympathetic and sympathetic innervation from the greater and deep petrosal nerves. The petrosal nerves
5
join to form the vidian nerve that enters the pterygopalatine ganglion, which sends branches to the zygomatic nerve that anastomose with the lacrimal nerve to reach the gland.
MIDDLE CRANIAL BASE Endocranial Surface The endocranial surface of the middle portion of the middle cranial base, formed by the sphenoid and temporal bones, has medial and lateral parts (Figs. 2, 3, 5, and 9). The medial part is formed by the body of the sphenoid bone, the site of the tuberculum sellae, pituitary fossa, middle and posterior clinoid processes, the carotid sulcus, and the dorsum sellae (Fig. 8). The lateral part is formed by the lesser and greater sphenoid wings, with the superior orbital fissure between them (Figs. 3 and 5). The lesser wing is connected to the body of the sphenoid bone by an anterior root, which forms the roof of the optic canal, and by a posterior root, also called the optic strut, which forms the floor of the optic canal and separates the optic canal from the superior orbital fissure (Fig. 3). The greater wing forms the largest part of the endocranial surface of the middle fossa, with the squamosal and the petrosal parts of the temporal bone completing this surface. The superior orbital fissure transmits the oculomotor, trochlear, ophthalmic, and abducens nerves, a recurrent meningeal artery, and the superior and inferior ophthalmic veins (6). The maxillary and mandibular nerves pass through the foramen rotundum and ovale, both located in the greater wing of the sphenoid. The not infrequently occurring sphenoidal emissary foramen, located anteromedial to the foramen spinosum, gives passage to a vein connecting the cavernous sinus and the pterygoid venous plexus. The upper surface of the petrous bone is grooved along the course of the greater and lesser petrosal nerves (Fig. 5) (7). The carotid canal extends upward and medially and provides passage to the internal carotid artery and carotid sympathetic nerves in their course to the cavernous sinus. The posterior trigeminal root reaches the middle fossa and the impression on the upper surface of the petrous bone where Meckel’s cave and the semilunar ganglion sit. The roof of the carotid canal opens below the trigeminal ganglion near the distal end of the carotid canal (Figs. 5, 6, and 9). The arcuate eminence approximates
←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− Figure 1 Anterior and middle cranial base. (A) On the left side, the floor of the anterior fossa and the upper portion of the maxilla have been removed to expose the structures deep to the anterior and middle cranial fossa. The frontal, ethmoidal, and sphenoid sinuses and the nasal cavity lie below the medial part of the anterior cranial base. The orbit and maxilla are located below the lateral part of the anterior cranial base. The sphenoid sinus and sella are located in the medial part of the middle cranial base, and the infratemporal and pterygopalatine fossa are located below the lateral part of the middle cranial base. The carotid arteries pass upward on the medial part of the middle cranial base and are intimately related to the sphenoid and cavernous sinuses. The infratemporal fossa, which contains branches of the mandibular nerve, pterygoid muscles, pterygoid venous plexus, and maxillary artery, is located below the middle cranial base and greater sphenoid wing. The alveolar process of the maxilla, which encloses the roots of the upper teeth, has been preserved on the left side. The maxillary nerve enters the pterygopalatine fossa, which is located medial to the infratemporal fossa between the posterior wall of the maxilla and the pterygoid process of the sphenoid bone. (B) Superior view of the anterior and middle cranial base. The infratemporal fossa is located posterolateral to the maxilla. The right ethmoid air cells are exposed on the medial side of the right orbit. The nasal cavity extends upward between the ethmoid sinuses. (C) Oblique anterior view. The facial structures on the right side have been removed to expose the orbital apex located above the maxillary sinus. The wall of the right maxillary sinus forms the floor of the orbit, much of the lateral wall of the nasal cavity, and the anterior wall of the pterygopalatine and infratemporal fossa. On the left side, the mandibular nerve enters the infratemporal fossa. The maxillary nerve enters the pterygopalatine fossa, which is located in the lateral wall of the nasal cavity and contains the maxillary nerve, pterygopalatine ganglion, and terminal branches of the maxillary artery. (D) Anterior view. The orbital apex is located above the pterygopalatine fossa. The frontal branch of the ophthalmic nerve passes along the roof of the orbit, and the infraorbital branch of the maxillary nerve courses in the floor of the orbit. The posterior ethmoid air cells are located medial to the orbital apex. The vomer forms the posterior part of the nasal septum and attaches to the maxilla and palatine bones below and to the body of the sphenoid bone above. The sphenoid sinus is located in the middle cranial base below the sella turcica. The upper brain stem is seen in the posterior part of the exposure (1).) Abbreviations: A., artery; Alv., alveolar; Br., branch; Car., carotid; Cart., cartilage; CN, cranial nerve; Eth., ethmoid; Eust., eustachian; Front., frontal; Gang., ganglion; Infraorb., infraorbital; Infratemp., infratemporal; M., muscle; Max., maxillary; N., nerve; Nasolac., nasolacrimal; Orb., orbital; Pit., pituitary; Pteryg., pterygoid; Pterygopal., pterygopalatine; Sphen., Sphenoid.
6
Martins and Rhoton
A Semicirc. canals Trautman’s triangle
Temp. dura Lat. Rec. M. CN III CN II CN IV
Vestibule Cochlea Jug. bulb V1
Gr. Pet. N.
Inf. Rec. M.
Car. N.
V3
V2
Infraorb. N.
Vidian N. Car. A.
Pterygopal. fossa Max. sinus Eust. tube Max. A.
Vert. A. Int. Jug. V. Infratemp. fossa
B Front. lobe Temp. lobe Ped.
Lat. Rec. M.
CN V
V1 Sphen. sinus
Gr. Pet. N. Cochlea
V2 Vidian N.
CN IX-XI
V3
Pterygopal. fossa Zygoma Eust. tube Max. A.
Car. A. Infratemp. fossa Figure 2 Lateral view of the anterior, middle, and posterior cranial base. (A) The bone and structures lateral to the orbit, infratemporal, and pterygopalatine fossa, and the parapharyngeal space and petrous part of the temporal bone have been removed to expose the structures below the anterior, middle, and posterior cranial base. The orbit and maxillary sinus are located below the anterior cranial base. The infratemporal and pterygopalatine fossae and the parapharyngeal space are located below the middle cranial base, and the suboccipital area is located below the temporal and occipital bones. The first trigeminal division is related to the upper part of the orbit. The second trigeminal branch is related to the lower part of the orbit and maxilla. The mandibular nerve exits the cranium through the foramen ovale and enters the infratemporal fossa. The pterygoid and levator and tensor veli palatini muscles have been removed to expose the eustachian tube and its opening into the nasopharynx. The lateral part of the temporal bone has been removed to expose the cochlea, vestibule, and semicircular canals. The petrous carotid passes upward and turns medially below the cochlea. The sigmoid sinus turns downward under the semicircular canals and vestibule where the jugular bulb is located. The segment of the vertebral artery passing behind the atlanto-occipital joint is located below the posterior cranial base. (B) The dura has been opened to show the relationships of the frontal and temporal lobes and the cerebellum to the cranial base. The orbit is exposed below the frontal lobe. The pterygopalatine and infratemporal fossae and the temporal bone are located below the temporal lobe. The jugular bulb and internal jugular vein have been removed to show cranial nerves IX through XII exiting the jugular foramen (1). Abbreviations: A., artery; Car., carotid; CN, cranial nerve; Eust., eustachian; Front., frontal; Gr., greater; Inf., inferior; Infraorb., infraorbital; Infratemp., infratemporal; Int., internal; Jug., jugular; Lat., lateral; M., muscle; Max., maxillary; N., nerve; Ped., peduncle; Pet., petrosal; Pterygopal., pterygopalatine; Rec., rectus; Semicirc., semicircular; Sphen., sphenoid; Temp., temporal; V., vein; Vert., vertebral.
Chapter 1: Anatomy of the Cranial Base
7
B
A
Crib. plate Front. bone
Eth. bone Sphen. ridge
Maxilla
Less. wing Vomer Optic canal Sphen. body For. lacerum For. ovale For. ovale Temp. bone Car. canal Temp. bone Jug. for. Pet. part
Gr. wing Ant. clin. For. spinosum Pet. Ridge
C
Palat. bone Horiz. plate Gr. wing Pteryg. proc. Mandib. fossa Clivus Occip. bone
D
Front. bone
Infraorb. for.
Supraorb. notch
Zygoma
Sup. Orb. Fiss. Maxilla Nasal bone Inf. Orb. Fiss.
Less. wing Gr. wing
Palat. bone Horiz. plate Vomer
Pteryg. proc. Mandib. fossa
Eth. Perp. plate Inf. concha
Vomer
Temp. bone Pet. part
Occip. bone
Jug. for.
Maxilla
Mandible
Figure 3 Osseous relationships of the anterior and middle cranial base. (A) On the endocranial surface, the anterior and middle cranial bases correspond to the anterior and middle fossae. The anterior part of the cranial base is separated from the middle fossa by the sphenoid ridge and the chiasmatic sulcus. The middle cranial base is separated from the posterior cranial base by the dorsum sellae and the petrous ridges. The upper surface of the anterior cranial base is formed by the frontal bone, which roofs the orbit; the ethmoid bone, which is interposed between the frontal bones and the site of the cribriform plate; and the lesser wing and anterior part of the body of the sphenoid, which forms the posterior part of the floor of the anterior fossa. The upper surface of the middle cranial base floor is formed by the greater sphenoid wing and posterior two-thirds of the sphenoid body anteriorly and the upper surface of the temporal bone posteriorly. The posterior part of the cranial base is formed by the temporal and occipital bones. The cribriform plate, sella, and clivus are located in the medial part of the cranial base. The lateral part of the cranial base is located above the orbits, pterygopalatine and infratemporal fossae, and the subtemporal and lateral part of the suboccipital areas. (B) Exocranial surface of the cranial base. This surface is more complicated than the endocranial surface. It is not demarcated into three well-defined fossae as is the endocranial surface. The exocranial surface is formed by the maxilla, zygomatic, palatine, sphenoid, temporal, and occipital bones, and the vomer. The maxilla, orbits, and nasal cavity are located below the anterior fossa. The anterior part of the hard palate is formed by the maxilla and the posterior part is formed by the palatine bone. The anterior part of the zygomatic arch is formed by the zygoma and the posterior part by the squamosal part of the temporal bone. The mandibular fossa on the lower surface of the temporal squama is located below the posterior part of the middle fossa. The vomer attaches to the lower part of the body of the sphenoid and forms the posterior part of the nasal septum. (C) Anterior view. The orbital rim is formed by the frontal bone, zygoma, and maxilla. The roof of the orbit is formed by the frontal and sphenoid bones; the lateral wall by the greater sphenoid wing and the zygomatic bone; the floor by the maxilla, except for a small part of the posterior floor formed by the palatine bone; and the medial wall of the orbit by the maxilla, lacrimal, and ethmoid bones. The nasal bone is interposed above the anterior nasal aperture between the maxillae. The nasal cavity is located between the ethmoid bones above and the maxillae and palatine bones, and sphenoid pterygoid process below. It is roofed by the frontal and ethmoid bones and the floor is formed by the maxillae and palatal bones. The osseous nasal septum is formed by the perpendicular ethmoid plate and the vomer. The inferior concha is a separate bone, and the middle and superior conchae are appendages of the ethmoid bone. The orbit opens through the superior orbital fissure into the middle fossa and through the inferior orbital fissure into the pterygopalatine and infratemporal fossae. (D) Anteroinferior view of the cranial base. The anterior part of the hard palate is formed by the maxillae and the posterior part is formed by the horizontal plate of the palatine bone. The vomer forms the posterior part of the nasal septum and divides the posterior nasal aperture in the midline. The infratemporal fossa is located below the greater sphenoid wing. The clivus is formed above by the body of the sphenoid bone and below by the basal part of the occipital bone. The petrous apex is interposed between the greater sphenoid wing and the clival part of the occipital bone. The mandibular condyles sit in the mandibular fossa located below the posterior part of the middle fossa on the inferior surface of the squamosal part of the temporal bone. (E) The cranial base is formed, in the lateral view, from anterior to posterior, by the maxilla and the frontal, zygomatic, sphenoid, temporal, and occipital bones. The zygomatic and frontal bones form the lateral part of the orbital rim. The pterion on the greater sphenoid wing marks the lateral end of the sphenoid ridge. The keyhole, a burr hole that exposes the dura of the anterior fossa and the periorbita in its depth, is located just above the frontozygomatic suture, behind the superior temporal line. The zygomatic arch is formed by the zygomatic bone and the squamosal part of the temporal bone. The condylar fossa, in which the mandibular condyle sits, is positioned above on the lower surface of the squamosal part of the temporal bone and posteriorly on the tympanic part of the temporal bone. The lower end of the pterygoid process unites with the posterior maxilla, but above, the process separates from the maxilla to create the pterygomaxillary fissure, which opens medially into the pterygopalatine fossa. (Continued).
8
Martins and Rhoton
E
F Keyhole Pterion
Temp. bone Squam. part Mandib. fossa Occip. bone
Zygoma Infraorb. for. Maxilla Pteryg. Proc. Mandib. cond. Max. sinus
Eth. perp. plate
Eth. Air Cells
Coronoid proc. Vomer Pterygopal. fossa
Pteryg. proc.
For. ovale
H
G Maxilla Front. bone
Supraorb. for.
Front. bone
Zygoma Eth. sinus Sup. orb. fiss. Less. sing
Eth. air cells Gr. wing Inf. orb. fiss. Vomer Vidian canal Pteryg. Proc. removed
Palat bone Gr. wing Horiz. plate Infratemp. fossa For. ovale Temp. bone For. spinosum
Clivus Occip. bone
For. rotundum Sphen. sinus Mandib. fossa
H
Figure 3 (Continued) (F) Inferior view of a cross section extending through the maxillae. The maxilla, which contains a large air-filled sinus, forms the anteromedial wall of the infratemporal fossa, the anterior wall of the pterygopalatine fossa, the lateral wall of the nasal cavity, the anterior portion of the hard palate, and much of the floor of the orbit. The pterygopalatine fossa is located between the pterygoid process and the posterior maxillary wall. The nasal septum is formed anteriorly and above by the perpendicular ethmoid plate and posteriorly and below by the vomer. (G) The right half of the maxilla and zygomatic arch has been removed. The inferior orbital fissure is located between the greater sphenoid wing and the maxilla. The right orbital roof and ethmoid air cells have been preserved. The right pterygoid process has been removed at its junction with the sphenoid body. The roof of the vidian canal, which extends through the base of the pterygoid process, has been preserved. (H) Anteroinferior view of the cranial base. The midline of the cranial base is formed, from anterior to posterior, by the frontal, ethmoid, sphenoid, and occipital bones. The roof of the orbit is formed by the frontal bone and lesser sphenoid wing. The ethmoidal sinuses are located anterior to the sphenoid sinus between the orbits. (I) Lateral view of the pterygomaxillary fissure. The pterygomaxillary fissure is located between the posterior maxillary wall and the pterygoid process. The pterygomaxillary fissure opens from the infratemporal fossa into the pterygopalatine fossa. The mandibular fossa is formed above by the squamosal part of the temporal bone and posteriorly by the tympanic part of the temporal bone, which also forms the anterior and lower wall of the external auditory meatus. (J) Anterior view through the maxillary sinus. The anterior and posterior walls of the maxillary sinus have been removed to expose the pterygoid process, which forms the posterior wall of the pterygopalatine fossa. The lower part of the superior orbital fissure is seen through the upper part of the maxillary sinus. The foramen rotundum opens into the pterygopalatine fossa and is separated from the superior orbital fissure by the maxillary strut. The vidian canal opens through the pterygoid process below and medial to the foramen rotundum. (K) Anterior view of a cranium sectioned through the posterior part of the ethmoid and maxillary sinuses. The ethmoidal sinuses are located anterior to the sphenoid body and sphenoid sinus. The part of the posterior wall of the maxilla forming the anterior wall of the pterygopalatine fossa has been preserved. The perpendicular plate of the palatine bone forms the medial wall of the pterygopalatine fossa. The ethmoidal sinus overlaps the lateral margin of the sphenoid ostia. The superior orbital fissure is located between the lesser and greater sphenoid wings and the sphenoid body. The infratemporal fossa is located below the greater wing of the sphenoid. The temporal fossa, which contains the temporalis muscle, is located between the greater wing and the zygomatic arch. (L) The posterior wall of the maxilla and ethmoidal sinuses have been removed to expose the sphenoid sinus and pterygopalatine fossa. The lateral wing of the sphenoid sinus extends laterally into the pterygoid process below the foramen rotundum. Septae divide the sphenoid sinus. The vidian canal opens through the base of the pterygoid process into the pterygopalatine fossa. (M) The osseous cross section has been extended posteriorly to just in front of the superior orbital fissure. The optic strut extends from the base of the anterior clinoid to the sphenoid body and separates the optic canal from the superior orbital fissure. The foramen rotundum is located below the medial part of the superior orbital fissure. The vidian canal opens into the pterygopalatine fossa below and medial to the foramen rotundum. (Continued).
Chapter 1: Anatomy of the Cranial Base
Gr. wing
I
9
J
Temp. bone Squam. part Mandib. fossa
Inf. orb. fiss. Zygoma
Pterygomax. fiss. Infratemp. fossa
Sup. orb. fiss.
For. rotundum
Maxilla Vidian canal Pteryg. proc. Pteryg. proc.
K Sup. orb. fiss. Gr. wing
Less. wing Sphen. ostia
Max. sinus Post. wall Ant. wall
L Gr. wing
Sphen. septa Sphen. sinus For. rotundum
Eth. sinus Temp. bone Squam. part
Max. strut
Infratemp. fossa Palat. bone Perp. plate
Vidian canal
Pteryg. proc.
Max. sinus Post. wall
M
Gr. waing
Optic canal
N
Ant. clin. Less. wing Optic strut Sup. orb. fiss. Sup. orb. fiss. For. rotundum Gr. wing Vidian canal Pteryg. proc.
Zygoma
Ant. clin. Optic canal Optic strut Sphen. body Vidian canal Pteryg. proc. Maxilla Med. pteryg. plate
Lat. pteryg. plate Palat. bone Horiz. plate
Figure 3 (Continued) (N) Posterior view of the specimen in K showing the anterior part of the middle fossa from behind. The superior orbital fissure is positioned below the lesser sphenoid wing. The optic strut extends from the base of the anterior clinoid to the sphenoid body and separates the optic canal from the superior orbital fissure. The greater wing extends laterally to form part of the floor and anterior and lateral walls of the middle fossa. The medial and lateral pterygoid plates project backward from the pterygoid process. The horizontal plate of the palatine bone forms the posterior part of the hard palate. The posterior opening into the vidian canal is located above the medial pterygoid plate and extends forward through the pterygoid process at its junction with the sphenoid body (1). Abbreviations: Ant., anterior; Car., carotid; Clin., clinoid; Cond., condyle; Crib., cribriform; Eth., ethmoid; Fiss., fissure; For., foramen; Front., frontal; Gr., greater; Horiz., horizontal; Inf., inferior; Infraorb., infraorbital; Infratemp., infratemporal; Jug., jugular; Lat., lateral; Less., lesser; Mandib., mandibular; Max., maxillary; Med., medial; Orb., orbital; Occip., occipital; Palat., palatine; Perp., perpendicular; Pet., petrosal, petrous; Post., posterior; Proc., process; Pteryg., pterygoid; Pterygomax., pterygomaxillary; Pterygopal., pterygopalatine; Sphen., sphenoid; Squam., squamosal; Sup., superior; Supraorb., supraorbital; Temp., temporal.
10
Martins and Rhoton
A
B
Front. sius Front. bone Eth. sinus Front. N. CN IV Sphen. ridge
Crib. plate
Olf. bulb
Eth. bone Olf. Tr.
Planum Optic sheath
Sphen. bone
Less. wing MCA
ACA
D
C Front. N.
Trochlea Lac. N. CN IV
Sup. ophth. V. Crib. plate
Sup. Obl. M. Lev. M.
Olf. Tr.
Front. N. Nasocil. N. Sup. rec. M. Sup. ophth. V.
Eth. sinus
Sphen. sinus
Optic sheath MCA CN II Ant. Clin. removed Car. A.
Figure 4 Anterior fossa, orbit, and perinasal sinuses. (A) Superior view. The anterior cranial fossa is formed by the frontal, ethmoid, and sphenoid bones. The frontal bone splits anteriorly into two laminae, which enclose the frontal sinus. The ethmoid bones, which contain the ethmoid air cells and are the site of the crista galli and cribriform plate, are interposed between the frontal bones. Posteriorly, the frontal and ethmoid bones join the sphenoid bone, which encloses the sphenoid sinus and has the pituitary fossa on its upper surface. The olfactory bulbs and tracts have been preserved. (B) The roof of the right orbit has been removed to expose the periorbita. The right anterior clinoid process and roof of the optic canal have been removed to expose the optic nerve enclosed within the optic sheath as it passes through the optic canal to reach the orbital apex. (C) The frontal, trochlear, and lacrimal nerves can be seen through the periorbita. The trochlear nerve crosses above the orbital apex to reach the superior oblique muscle. (D) The orbital fat has been removed and the sphenoid sinus opened. The frontal branch of the ophthalmic nerve courses above the levator muscle. The ophthalmic artery, nasociliary nerve, and superior ophthalmic vein are located medially in the anterior part of the orbit and cross between the optic nerve and the superior rectus muscle and are thus situated on the lateral side of the optic nerve at the orbital apex. (E) Enlarged view. The superior oblique muscle has been retracted medially to expose the anterior and posterior ethmoidal branches of the ophthalmic artery and nasociliary nerve entering the anterior and posterior ethmoidal canal. The trochlea of the superior oblique muscle is attached to the superomedial margin of the orbit just behind the orbital rim. The frontal nerve divides into supraorbital and supratrochlear branches. (F) The levator and superior rectus muscle have been retracted posteriorly to expose the nasociliary nerve, ophthalmic artery, and superior ophthalmic vein passing above the optic nerve. (G) Superior view of the anterior fossa in another specimen. The nasal cavity, sphenoid sinus, and orbit have been unroofed. The dura has been removed from the roof and lateral wall of the cavernous sinus. The medial strip below the anterior cranial base is formed, from anterior to posterior, by the frontal, ethmoidal, and sphenoid sinuses. The orbital fat has been removed to expose the intraorbital structures. The frontal nerve courses above the levator muscle. The trochlear nerve passes above the annular tendon to reach the superior oblique muscle. The trochlea of the superior oblique muscle is attached in the superomedial part of the anterior orbit. The lacrimal nerve courses above the lateral rectus muscle. The ophthalmic artery and superior ophthalmic vein are seen in the interval between the levator and superior oblique muscle. The anterior and posterior ethmoidal branches of the ophthalmic artery course through the anterior and posterior ethmoidal canals. (Continued).
Chapter 1: Anatomy of the Cranial Base
E
11
F
Supratroch. N.
Supraorb. N.
Trochlea Sup. Ophth. V. Ant. Eth. A. & N.
Ophth. A.
Nasocil. N.
Front. N. Sup. Obl. M.
Post. Eth. A. & N.
Med. Rec. M. Sup. Obl. M.
Sup. Rec. M. Ophth. A.
Lev. M.
Sup. Ophth. V. Nasocil. N.
Sup. Ophth. V. Lev. M.
Sup. Rec. M.
G
H
Nasocil. N.
Trochlea Lev. M. Sup. Obl. M. Lev. M. CN IV Sup. Rec. M. Sup. Rec. M.
Sup. Obl. M. Ophth. A. Ant. Eth. A.
Clin. Seg. Sup. Ophth. V.
CN II
CN III
Anular tendon
CN II Front. N. Lac. N. Ophth. A.
Post. Eth. A. CN IV Ophth. A.
Car. A.
Ophth. A.
V2 V1
V2 CN III Cav. Seg. CN IV
V1 V3
Figure 4 (Continued) (H) Enlarged view of cavernous sinus, superior orbital fissure, and orbital apex. The superior oblique, levator, and superior rectus muscles have been removed. The ophthalmic artery and nasociliary nerve enter the orbital apex on the lateral side of the optic nerve and cross between the optic nerve and superior rectus muscle to reach the medial part of the orbit. The optic nerve has been elevated to expose the ophthalmic artery, which courses through the optic canal on the lower side of the optic nerve and enters the orbital apex on the lateral side of the optic nerve. The ophthalmic artery then crosses medially between the optic nerve and the superior rectus muscle, as does the nasociliary nerve. The maxillary nerve exits the foramen rotundum to enter the pterygopalatine fossa, and the mandibular nerve exits the foramen ovale to enter the infratemporal fossa (1). Abbreviations: A., artery; ACA, anterior cerebral artery; Ant., anterior; Car., carotid; Cav., cavernous; Clin., clinoid; CN, cranial nerve; Crib., cribriform; Eth., ethmoid, ethmoidal; Front., frontal; Lac., lacrimal; Less., lesser; Lev., levator; M., muscle; Med., medial; MCA, middle cerebral artery; N., nerve; Nasocil., nasociliary; Obl., oblique; Olf., olfactory; Ophth., ophthalmic; Post., posterior; Rec., rectus; Seg., segment; Sphen., sphenoid; Sup., superior; Supraorb., supraorbital; Supratroch., supratrochlear; Tr., tract; V., Vein.
12
Martins and Rhoton
A
B Temp. M. Temp. M. Max. A.
CN II Ophth. A.
V2 Lat. Pteryg. M.
Gr. wing
V2 V1 CN III Cav. Seg. CN IV
CN III CN IV V3
Mandib. Cond.
V3
Pteryg. Plex.
CN VI
CN VI
Eust. Tube
Pet. Seg. Gr. Pet. N. CN VII Mast. antrum CN VIII CN IX-XI
CN VII CN VIII
C Lat. Pteryg. M. Mandib. Cond.
Max. A.
Ext. Ac. meatus V2
Chorda Tymp. N. Gr. Pet. N. Eust. tube Cochlea Semicirc. canal
V3 V1 Pet. Seg. CN III CN IV
Mast. antrum
D V1 V2 V3 Vidian N.
Petroling. Lig. Cav. Seg. CN VI CN IV CN III Petrosphen. Lig.
Chorda Tymp. N. Less. Pet. N. Deep Pet. N. Gr. Pet. N.
Pet. Seg. CN VII
CN VIII
Figure 5 Superior view of middle cranial base. (A) The floor of the middle fossa has been preserved. The anterior part of the floor of the middle fossa is formed by the greater sphenoid wing, which roofs the infratemporal fossa, and the posterior part of the floor is formed by the upper surface of the temporal bone. The internal acoustic meatus, mastoid antrum, and tympanic cavities have been unroofed. The dural roof and lateral wall of the cavernous sinus have been removed. The petrous segment of the internal carotid artery is exposed lateral to the trigeminal nerve. The temporalis muscle is exposed in the temporal fossa lateral to the greater sphenoid wing. (B) The floor of the middle fossa has been removed to show the relationship below the floor. The temporalis muscle descends medial to the zygomatic arch in the temporal fossa to insert on the coronoid process of the mandible. The infratemporal fossa is located medial to the temporal fossa, below the greater sphenoid wing, and contains the pterygoid muscles and venous plexus and branches of the mandibular nerve and maxillary artery. The mandibular condyle is located below the posterior part of the middle fossa floor, which is formed by the temporal bone. (C) Enlarged view of the posterior part of the area below the middle fossa floor. The roof of the temporal bone, which forms the posterior part of the floor of the middle fossa, has been opened to expose the mastoid antrum, eustachian tube, semicircular canals, cochlea, the nerves in the internal acoustic meatus, and the mandibular condyle. (D) The trigeminal nerve has been reflected forward. The abducens nerve passes below the petrosphenoid ligament and through Dorello’s canal. The petrous segment of the carotid passes below the petrolingual ligament to enter the cavernous sinus. The greater petrosal nerve is joined by the deep petrosal branch of the carotid sympathetic plexus to form the vidian nerve, which passes forward in the vidian canal, which has been unroofed. The lesser petrosal nerve arises from the tympanic branch of the glossopharyngeal nerve, which passes across the promontory in the tympanic nerve plexus and regroups to cross the floor of the middle fossa, exiting the cranium to provide parasympathetic innervation through the otic ganglion to the parotid gland. The tensor tympani muscle and eustachian are layered, with the former above the latter, along and separated from the anterior surface of the petrous carotid by a thin layer of bone (1). Abbreviations: A., artery; Ac., acoustic; Cav., cavernous; CN, cranial nerve; Cond., condyle; Eust., eustachian; Ext., external; Gr., greater; Lat., lateral; Less., lesser; Lig., ligament; M., muscle; Mandib., mandibular; Mast., mastoid; Max., maxillary; N., nerve; Ophth., ophthalmic; Petroling., petrolingual; Pet., petrosal, petrous; Petrosphen., petrosphenoid; Plex., plexus; Pteryg., pterygoid; Seg., segment; Semicirc., semicircular; Temp., temporalis; Tymp., tympani.
Chapter 1: Anatomy of the Cranial Base
B
13
Crib. Plate
Nasolac. duct
A
Maxilla Max. sinus Palat. Bone
Inf. meatus
Eth. sinus
Vome r
Inf. Orb. Fiss. Vidian canal
Pteryg. Proc . Gr. Wing Gr. wing For. Oval e For. Lacerum Pterygopal. fossa Sulc. Tuba e
Pteryg. Proc. removed Car. canal
Vomer Palat. bone Perp. palate Pteryg. Proc.
Infratemp. fossa For. ovale
C
D
Max. A.
Max. sinus
Mid. concha Pteryg. Plex.
Infraorb. N. Temp. M.
Lat. Pteryg. M.
Max. sinus
Mandible Cond. Proc.
Pterygopal. fossa
Max. A.
Eust. Tube Tens. Vel. Pal. M. Lev. Vel. Pal. M.
Pteryg. Proc. Infratemp. fossa Mandible Max. A. Car. A. Parotid Gl. Int. Jug. V.
Car. Sheath Eust. tube Styloid Proc. Rosenmueller’s faossa Clivus Parotid Gl. Int. Jug. V.
Long. Cap. M.
Car. A.
CN IX-XII
Rec. Cap. Lat. M.
Figure 6 (A) Inferior view of cranial base. The right pterygoid process has been sectioned and removed at its junction with the greater wing and body of the sphenoid bone to expose the pterygopalatine fossa and the vidian canal. The vidian nerve, formed by the union of the superficial and deep petrosal nerves, courses in the vidian canal, which passes through the root of the pterygoid process. It opens posteriorly at the anterolateral margin of the foramen lacerum and anteriorly into the medial portion of the pterygopalatine fossa. The sulcus tubae, which is the attachment site of the cartilaginous part of the eustachian tube to the cranial base, is located on the extracranial surface of the sphenopetrosal fissure, anterolateral to the foramen lacerum and the carotid canal, and posteromedial to the foramina ovale and spinosum. The lateral part of the inferior orbital fissure opens into the infratemporal fossa, located below the greater sphenoid wing, and the medial part opens into the pterygopalatine fossa, located below the orbital apex between the maxilla and pterygoid process. The right zygomatic arch has been removed. (B) Inferior view of axial section of a cranium at the level of the maxillary sinus. The pterygopalatine fossa is located between the posterior wall of the maxillary sinus and the pterygoid process. The roof of the maxillary sinus forms the floor of the orbit. The infratemporal fossa is located below the greater wing of the sphenoid and opens medially into the pterygopalatine fossa. The medial wall of the pterygopalatine fossa is formed by the perpendicular plate of the palatine bone, which has an opening, the sphenopalatine foramen, through which branches of the maxillary artery and nerve reach the nasal cavity. The ethmoid air cells are located medial to the orbit. (Continued).
14
Martins and Rhoton
Max. sinus
E Max. A.
F
Max. sinus
Pterygopal. fossa Max. A.
Lat. Pteryg. M. Pteryg. Proc. Max. A. V3
Gr. wing
Eust. tube
Pterygopal. fossa Palat. bone Perp. plate Pteryg. Proc.
Infratemp. fossa Rosenmueller’s fossa Vomer
Mandible
V3
H
G Infraorb. N. Zygo. N.
Infraorb. N. V2
Pterygopal. Gang.
V2 V3
Lat. Pteryg. M.
Eust. tube
For. rotundum Vidian N. V3 For. ovale Eust. tube
Sphen. sinus Sella Pet. Seg. Clivus
Styloid Proc. Pet. Seg. CN VII For. lacerum Int. Jug. V.
CN IX-XII Sig. sinus
Figure 6 (Continued) (C) Inferior views of an axial section of the cranial base. The infratemporal fossa is surrounded by the maxillary sinus anteriorly, the mandible laterally, the pterygoid process anteromedially, and the parapharyngeal space posteromedially. It contains the mandibular nerve and maxillary artery and their branches, the medial and lateral pterygoid muscles, and the pterygoid venous plexus. The posterior nasopharyngeal wall is separated from the lower clivus by the longus capitis, and the nasopharyngeal roof rests against the upper clivus and floor of the sphenoid sinus. (D) Enlarged view with highlighting of the pre- (red) and poststyloid (yellow) compartments of the parapharyngeal space. The styloid diaphragm, formed by the anterior part of the carotid sheath, separates the parapharyngeal space into pre- and poststyloid parts. The prestyloid compartment, a narrow fat-containing space between the medial pterygoid and tensor veli palatini muscle, separates the infratemporal fossa from the medially located lateral nasopharyngeal region containing the tensor and levator veli palatini and the eustachian tube. The poststyloid compartment, located behind the prestyloid part, contains the internal carotid artery, internal jugular vein, and the cranial nerves IX through XII. (E) Some of the lateral pterygoid muscle has been removed to expose the branches of the mandibular nerve in the infratemporal fossa. The lower part of the pterygoid process has been removed to expose the maxillary artery in the pterygopalatine fossa. The pharyngeal recess (fossa of Rosenm¨uller) projects laterally from the posterolateral corner of the nasopharynx below the foramen lacerum. (F) Enlarged view. The pterygopalatine fossa is located between the posterior maxillary wall anteriorly, the sphenoid pterygoid process posteriorly, the perpendicular plate of the palatine bone medially, and the infratemporal fossa laterally. The medial part of the eustachian tube has been removed. (G) The pterygoid process has been removed to expose the maxillary nerve passing through the foramen rotundum to enter the pterygopalatine fossa where it gives rise to the infraorbital and zygomatic nerves and communicating rami to the pterygopalatine ganglion. The vidian nerve exits the vidian canal and joins the pterygopalatine ganglion. The terminal part of the petrous carotid is exposed above the foramen lacerum. (H) Enlarged view of the region of the carotid canal and jugular foramen. The bone below the carotid canal has been removed to expose the petrous carotid. The deep portion of the parotid gland has been removed to expose the facial nerve at the styloid foramen. The sigmoid sinus hooks downward from the posterior fossa and opens into the internal jugular vein. A portion of the occipital condyle has been removed to expose the hypoglossal nerve joining the nerves exiting the jugular foramen to pass downward in the carotid sheath. The styloid process and facial nerve at the stylomastoid foramen are located on the lateral side of the internal jugular vein. The right half of the floor of the sphenoid sinus has been removed to expose the sella (1). Abbreviations: A., artery; Cap., capitis; Car., carotid; CN, cranial nerve; Cond., condyle; Crib., cribriform; Eth., ethmoid; Eust., eustachian; Fiss., fissure; For., foramen; Gang., ganglion; Gl., gland; Gr., greater; Inf., inferior; Infraorb., infraorbital; Infratemp., infratemporal; Int., internal; Jug., jugular; Lat., lateral, lateralis; Lev., levator; Long., longus; M., muscle; .Max., maxillary; Mid., middle; N., nerve; Nasolac., nasolacrimal; Orb., orbital; Pal., palatini; Palat., palatine; Perp., perpendicular; Pet., petrosal, petrous; Plex., plexus; Proc., process; Pteryg., pterygoid; Pterygopal., pterygopalatine; Rec., rectus; Seg., segment; Sig., sigmoid; Sphen., sphenoid; Sulc., sulcus; Temp., temporalis; Tens., tensor; V., vein; Vel., veli; Zygo., Zygomatic.
the position of the superior semicircular canal. A thin lamina of bone, the tegmen tympani, roofs the area above the middle ear and auditory ossicles on the anterolateral side of the arcuate eminence. The internal auditory canal can be identified below the floor of the middle fossa by drilling along a line
approximately 60 degrees medial to the arcuate eminence, near the middle portion of the angle between the greater petrosal nerve and arcuate eminence (Fig. 5). The petrous apex, medial to the internal acoustic meatus, is free of important structures.
Chapter 1: Anatomy of the Cranial Base
15
B
A
Front. N.
Olf. bulb Crib. plate
Sup. Ophth. V. Nasocil. N.
Lev. M. Front. N. Sup. Obl. M. Lac. N. CN IV
Ophth. A. Optic sheath CN IV Ant. Clin. removed
Optic canal
Ant. Clin. removed
Car. A.
CN II MCA ACA
C
D
Anular tendon Orb. apex
Max. Sinus Nasolac. duct Infraorb. N.
Optic Strut Falc. Lig.
Zygo. N.
Planum Optic sheath Clin. Seg.
Pit. Stalk
Orb. apex Optic sheath
Lam. Term.
Figure 7 Superior view of the anterior cranial base. (A) Both orbits have been unroofed to expose the periorbita. The optic canals have been unroofed and the anterior clinoids removed to expose the optic nerves, which are enclosed in the optic sheath within the optic canal. The frontal, trochlear, and lacrimal nerves can be seen through the periorbita. The roof of the ethmoidal sinuses and the olfactory bulbs sitting on the cribriform plate has been preserved. The anterior cerebral arteries course above the optic chiasm. (B) The intraorbital fat has been removed and the levator and superior rectus muscles have been retracted laterally to expose both globes, ophthalmic arteries, superior ophthalmic veins, and nasociliary nerves. (C) The orbital contents have been removed to expose the lateral wall and floor of the orbit. The maxillary sinuses are exposed below the orbital floors. The maxillary nerves give rise to the infraorbital nerve, which courses along the floor of the orbit to reach the cheek, and the zygomatic nerve, which courses along the lateral wall of the orbit to reach the malar eminence and temple. (D) Enlarged view. The optic nerves are enclosed within the optic sheath as they course through the optic canal. The annular tendon, from which the rectus muscles arise, surrounds the optic nerve and medial portion of the superior orbital fissure. Removal of the anterior clinoid exposes the clinoid segment of the carotid artery. The optic strut, which separates the optic canal and superior orbital fissure, has also been removed. The segment of anterior cerebral arteries passing above the chiasm has been removed to expose the lamina terminalis. The falciform dural fold extends across the optic nerve at the entrance into the optic canal (1). Abbreviations: A., artery; ACA, anterior cerebral artery; Ant., anterior; Car., carotid; Clin., clinoid; CN, cranial nerve; Crib., cribriform; Falc., falciform; Front., frontal; Infraorb., infraorbital; Lac., lacrimal; Lam., lamina; Lev., levator; Lig., ligament; M., muscle; Max., maxillary; MCA, middle cerebral artery; N., nerve; Nasocil., nasociliary; Nasolac., nasolacrimal; Obl., oblique; Olf., olfactory; Ophth., ophthalmic; Orb., orbital; Pit., pituitary; Seg., segment; Sup., superior; Term., terminalis; Zygo., zygomatic.
The middle cranial base can be divided into a lateral portion, containing the middle cranial fossa and the upper surface of the temporal bone, and a medial portion, the sellar and the parasellar region, where the pituitary gland and cavernous sinus are located (Figs. 3 and 8) (8). The basal temporal lobe, formed by the parahippocampal, occipitotemporal, infratemporal gyri, and uncus and supplied by branches of the anterior choroidal, posterior cerebral, and middle cerebral arteries, rests on the middle fossa floor. The cavernous sinus, situated between two layers of dura, is formed by an outer layer facing the brain, and inner or periosteum layer, covering the bone of the middle fossa (9). The inner layer splits into two parts when it reaches the cavernous sinus; one invests the
nerves and forms the inner layer of the lateral wall, and the medial layer faces the sphenoid body and forms the medial wall of the sinus. The same inner layer invests the oculomotor, trochlear, and ophthalmic nerves and the distal part of the abducens nerve in their course through the lateral wall of the cavernous sinus. The internal carotid artery with its vertical posterior bend, horizontal anterior bend, and clinoidal segments runs inside the cavernous sinus. The clinoidal segment of the internal carotid artery is between the distal and proximal dural rings and is covered by a layer of dura, which forms a collar, the carotid collar, around the artery (10). In a previous study, we found that the venous plexus, forming the cavernous sinus, extends through the lower
16
Martins and Rhoton
B
A
CN VI Crib. plate Sphenoeth. Rec. Sup. concha Mid. concha Inf. concha
Sup. Orb. Fiss. Sphen. sinus Sup. concha Int. Car. A. Mid. concha Eust. tube
V2 Cav. sinus
Inf. concha
D
C Front. sinus Crib. plate
Periorbita
Front. sinus ostium Max. sinus ostium
CN II Chiasm
Olf. bulb
Rosenmueller’s fossa
Infraorb. N. Max. Sinus Max. A.
Sup. Orb. Fiss. CN VI V2 Vidian N. Pterygopal. Fossa
F
E
CN VI Zygo. N. Infraorb. N.
Sup. Obl. M. Med. Rec. M.
Sphenopal. A. Sphenopal. Gang. Vidian N.
Anular tendon Max. A.
Inf. Rec. M. Gr. Palat. N.
Figure 8 Structures below the medial part of the anterior and middle cranial fossae. (A) Midsagittal section of the anterior and middle cranial base to the right of the nasal septum. The area below the medial part of the anterior cranial fossa is formed by the frontal and ethmoidal sinuses and the nasal cavity. The nasal cavity is divided into the inferior, middle, and superior meati and the sphenoethmoidal recess by the inferior, middle, and superior conchae. The inferior meatus is located below the inferior turbinate, and the sphenoethmoidal recess, into which the sphenoid sinus opens, is located above the superior turbinate. The central part of the middle cranial base is formed by the body of the sphenoid bone, which contains the sphenoid sinus and sella with the pituitary gland. The cribriform plate is located in the roof of the nasal cavity. The nasopharynx and the opening of the eustachian tube are located below the sphenoid sinus. (B) Some of the mucosa has been removed from the concha. The inferior concha is a separate bone attached to the maxilla. The middle and superior concha are appendages of the ethmoid bone. The carotid artery courses along the lateral margin of the sphenoid sinus. The prominence within the sphenoid sinus, formed by the superior orbital fissure, is located anterior to the intracavernous carotid, and the prominence overlying the maxillary nerve is located below the intracavernous carotid. (C) The middle and superior turbinates have been removed to expose the ostia of the maxillary and frontal sinuses. Both open into the middle meatus below the middle turbinate. The nasolacrimal duct opens below the inferior concha. Rosenm¨uller’s fossa is located behind the eustachian tube. (D) The medial wall of the maxillary sinus and the ethmoid air cells have been removed to expose the orbit. The optic nerve enters the orbit above the superior orbital fissure. The maxillary nerve exits the foramen rotundum to enter the pterygopalatine fossa. The vidian nerve passes through the vidian canal and enters the posterior margin of the sphenopalatine ganglion in the pterygopalatine fossa. The floor of the anterior cranial fossa forms much of the roof of the orbit and maxillary sinus forms most of the floor of the orbit. The abducens nerve is seen below the intracavernous segment of the internal carotid artery. The pterygopalatine fossa is located anterior to the sphenoid sinus and below the orbital apex. (E) The intraorbital fat has been removed to expose the superior oblique and medial and inferior rectus muscles. (F) Enlarged view of the pterygopalatine fossa. The maxillary nerve exits the foramen rotundum to enter the pterygopalatine fossa, where it gives rise to the infraorbital, zygomatic, and palatine nerves and communicating rami to the pterygopalatine ganglion. The vidian nerve exits the vidian canal to enter the pterygopalatine ganglion. The pterygopalatine fossa contains branches of the maxillary nerve, the junction of the vidian nerve with the pterygopalatine ganglion, and terminal branches of the maxillary artery (1). Abbreviations: A., artery; Car., carotid; Cav., cavernous; Crib., cribriform; CN, cranial nerve; Eust., eustachian; Fiss., fissure; Front., frontal; Gang., ganglion; Inf., inferior; Infraorb., infraorbital; Int., internal; M., muscle; Max., maxillary; Med., medial; Mid., middle; N., nerve; Obl., oblique; Olf., olfactory; Orb., orbital; Palat., palatine; Pterygopal., pterygopalatine; Rec., recess, rectus; Sphen., sphenoid; Sphenoeth., sphenoethmoidal; Sphenopal., sphenopalatine; Sup., superior; Zygo., zygomatic.
Chapter 1: Anatomy of the Cranial Base
A
17
B Sup. Temp. A.
Orb. Oculi M. Orb. Oculi M. CN VII Frontotemp. Plex. Parotid duct Parotid gland Mass. M. Parotid duct
C
Front. M.
CN VII Frontotemp. Plex.
D
E Temp. M. Temp. M.
Orb. Oculi M.
CN VII to Front. M.
Sup. Temp. A.
Sup. Temp. A. Zygomatic M.
Orb. Oris M.
Zygo. arch Sup. Temp. A.
CN VII Brs.
TM joint
Parotid duct Mass. M. CN VII Brs. Parotid gland Bucc. M.
Zygoma Mandib. Cond.
CN VII
Coronoid Proc.
Mass. M. Bucc. M. Bucc. M.
Platysma M.
Figure 9 (A) The branches of the facial nerve, which form a fine plexus in the fat pad overlying the temporalis fascia and are directed to the orbicularis oculi and frontalis muscle, have been dissected free and a small piece of black material placed deep to their fine branches to highlight this neural network in the fat pad. (B) Enlarged view of the facial nerve plexus innervating the orbicularis oculi and frontalis muscle. (C) Lateral view of the structures superficial to the anterior and middle cranial base. The frontotemporal and zygomatic branches of the facial nerve are exposed anterior to the parotid gland. The orbicularis oculi surrounds the orbit, and the frontalis muscle extends upward from the superior orbital rim. The levators of the lip and zygomaticus muscles are located in front of the maxilla. The orbicularis oris surrounds the mouth and the buccinator muscle surrounds the oral cavity deep to the masseter muscle. The parotid duct crosses the masseter muscle. The superficial temporal artery divides into anterior and posterior branches. The parotid gland has been removed to show the branches of the facial nerve. (D) The parotid gland has been removed to expose the facial nerve exiting the stylomastoid foramen. The facial nerve branch to the frontalis muscle has been preserved in the dissection and has been laid back against the temporalis muscle to show it crossing the zygomatic arch in its course to the forehead. The superficial temporal artery passes deep to the facial nerve in front of the ear. (E) The masseter muscle has been removed to expose the temporalis muscle inserting on the coronoid process. The buccinator muscle, which surrounds the oral cavity, is situated on the deep side of the masseter muscle. (F) The coronoid process and lower part of the temporalis muscle have been removed to expose the deep temporal branches of both the maxillary artery and the mandibular nerve passing upward along the greater sphenoid wing and temporal squama to enter the deep side of the temporalis muscle. The lateral pterygoid muscles extend backward from the pterygoid process and greater wing of the sphenoid to insert along the mandibular condyle and temporomandibular joint. (G) A craniotomy has been done to expose the floor of the middle fossa, and the lateral wall of the orbit has been removed to expose the extraocular muscles. The mandibular condyle has been removed and the pterygoid muscles reflected to expose the mandibular nerve at the foramen ovale. The pterygopalatine fossa is located behind the maxilla. The floor of the orbit and the upper part of the maxilla has been removed to expose the nasal cavity. (Continued).
18
Martins and Rhoton
F
G
Temp. M.
Front. lobe
Sup. Temp. A. Deep Temp. A. & N. Temp. lobe Lat. Rec. M.
Lat. Pteryg. M. Inf. Obl. M. Mandib. Cond.
Infraorbital N. Pterygopal. fossa
Mid. fossa floor Mandib. fossa Max. A. V3
Maxilla Inf. concha
Med. Pteryg. M.
H
I V2 Orbit V1
Pet. Seg.
Tymp. Memb. Rec. Cap. Ant. M. Long. Cap. M.
V2 Chorda Tymp. N. Gr. Pet. N. V3
V3
CN VII
Car. A.
Vert. A.
Eust. tube
Int. Jug. V. CN XI
CN VII
J
K Sig. sinus
Sup. Pet. sinus Sup. Pet. Sinus V1 Car. canal
Sig. sinus Jug. Bulb
V2
CN V Post. root
V3 Pons CN IX-XII
Car. A.
Car. A. Vert. A.
Figure 9 (Continued) (H) Enlarged view following resection of the floor of the middle fossa and the external auditory canal to expose the tympanic membrane and the mandibular nerve below the foramen ovale. The mastoid segment of the facial nerve has been preserved. The greater petrosal nerve crosses above the petrous carotid. The tensor tympani muscle and eustachian tube are layered along the anterior margin of the petrous carotid. (I) The eustachian tube and tensor tympani have been resected to expose the upper cervical and petrous carotid. The nasopharyngeal mucosa has been opened to expose the longus capitus and rectus capitus anterior muscles. (J) The carotid artery has been reflected forward out of the carotid canal. This exposes the petrous apex in front of the jugular foramen on the medial side of the internal carotid artery. (K) The petrous apex has been drilled and the dura opened below the trigeminal nerve to expose the upper anterior part of the posterior cranial fossa. A segment of the internal jugular vein and jugular bulb has been resected to expose the IX through XII cranial nerves below the jugular foramen and hypoglossal canal (1). Abbreviations: A., artery; Ant., anterior; Brs., branches; Bucc., buccinator; Cap., capitis; Car., carotid; CN, cranial nerve; Cond., condyle; Eust., eustachian; Front., frontal, frontalis; Frontotemp., frontotemporal; Gr., greater; Inf., inferior; Int., internal; Jug., jugular; Lat., lateral; Long., longus; M., muscle; Mandib., mandibular; Mass., masseter; Max., maxillary; Med., medial; Memb., membrane; N., nerve; Obl., oblique; Orb., orbital; Pet., petrosal, petrous; Plex., plexus; Post., posterior; Proc., process; Pteryg., pterygoid; Pterygopal., pterygopalatine; Rec., rectus; Seg., segment; Sig., sigmoid; Sup., superior; Temp., temporal, temporalis; Tymp., tympani, tympanic; V., vein; TM, temporomandibular; Vert., vertebral; Zygo., zygomatic.
Chapter 1: Anatomy of the Cranial Base
19
B
A
Ophth. A. Eth. sinus CN II
Orbit Med. Rec. M. Eth. sinus Sphen. ostia
Lat. Rec. M .
Septum Max. A.
Mid. concha Mid. meatus Mid. concha
Max. sinus
Inf. concha
Inf. meatus
Max. sinus
Inf. meatus
CN II Ophth. A.
D
C
Cav. Seg.
CN III CN II V1
Ophth. A. Cav. Seg.
CN II
Med. Rec. M.
CN VI
CN VI V2
Lat. Rec. M. Cav. sinus
Sphen. sinus
For. Rotundum
Inf. Rec. M. Pet. Seg. Infraorb. N.
Pet. Seg.
Pterygopal. fossa
Vidian N. Pterygopal. Gang. Max. A. Infratemp. fossa
Pterygopal. fossa Max. A.
Gr. Palat. N.
Max. sinus Pteryg. Proc. Eust. tube
Infratemp. fossa Lat. Pteryg. M.
Figure 10 (A) Anterior view of a coronal section, anterior to the sphenoid sinus, through the nasal cavity, orbits, and maxillary sinuses. The upper part of the nasal cavity is separated from the orbits by the ethmoidal sinuses. The lower part of the nasal cavity is bounded laterally by the maxillary sinuses. The middle concha projects medially from the lateral nasal wall at the junction of the roof of the maxillary and ethmoidal sinuses. The posterior ethmoid air cells are located in front of the lateral part of the sphenoid sinus. (B) The middle and inferior nasal conchae on the left side and the nasal septum and the posterior ethmoidal sinuses on both sides have been removed to expose the posterior nasopharyngeal wall, the anterior aspect of the sphenoid body, and the sphenoid ostia. The posterior ethmoid air cells overlap the lateral margin of the sphenoid ostia. (C) Enlarged view showing the relationships of the nasal cavity, pterygopalatine and infratemporal fossae, orbit, and sphenoid sinus. The nasopharynx is located below the sphenoid sinus. The pterygopalatine fossa is located in the lateral wall of the nasal cavity behind the upper part of the maxillary sinus and below the orbital apex. The posterior maxillary wall is so thin that the maxillary artery coursing in the pterygopalatine fossa can be seen through the bone. The sphenopalatine branch of the maxillary artery passes through the sphenopalatine foramen to reach the walls of the nasal cavity and the sphenoid face. (D) The posterior wall of the maxillary sinus has been removed to expose the pterygopalatine and infratemporal fossae and the internal carotid artery and nerves coursing through the cavernous sinus. The maxillary artery passes through the infratemporal fossa and enters the pterygopalatine fossa, where it gives rise to branches that follow the branches of the maxillary nerve. Some of these arteries course along the sphenoid face where careful hemostasis during transsphenoidal surgery reduces the need for nasal packing after transsphenoidal operations. The maxillary nerve exits the foramen rotundum to enter the pterygopalatine fossa where it gives rise to the infraorbital and greater palatine nerves and communicating rami to the pterygopalatine ganglion. The eustachian tube opens into the nasopharynx along the posterior edge of the medial pterygoid plate (1). Abbreviations: A., artery; Cav., cavernous; CN, cranial nerve; Eth., ethmoid; Eust., eustachian; For., foramen; Gang., ganglion; Gr., greater; Inf., inferior; Infraorb., infraorbital; Infratemp., infratemporal; Lat., lateral; M., muscle; Max., maxillary; Med., medial; Mid., middle; N., nerve; Ophth., ophthalmic; Palat., palatine; Pet., petrosal; Proc., process; Pteryg., pterygoid; Pterygopal., pterygopalatine; Rec., rectus; Seg., segment; Sphen., sphenoid.
20
Martins and Rhoton
A
B
Sphen. sinus
Sphen. sinus
Car. A. Cav. sinus
Infratemp. fossa Lat. Pteryg. M.
Clivus
Med. Pteryg. M.
V3 Max. A.
Nasopharynx Car. A. Palate
For. magnum
Mass. M. Oropharynx
D
C
Bas. A. Eust. tube V3
Lat. Pteryg. M.
V3 Car. A.
Vert. A. Alar Lig.
Max. A. Long. Cap. M. Car. A.
Atlas
Int. Jug. V. Car. A.
Mandib. Cond. Trans. Lig.
Odontoid
Styloid M. Vert. A. Long. Colli M.
E
F Eth. sinus
Infraorb. N. V2
Pterygopal. fossa
Infraorb. N.
V2 Pterygopal. Gang.
Sup. concha Comm. rami Max. A. Pterygopal. Gang.
Mid. concha
Gr. Palat. N. & A.
Inf. concha
Vidian N. Sphenopal. A.
Max. A.
Gr. Palat. N. & A.
Figure 11 Anterior view. Stepwise dissection of a cross section showing the relationships below the middle cranial base. (A) The soft palate, which has been preserved, is located at the level of the foramen magnum. The infratemporal fossa, located below the greater sphenoid wing and middle cranial fossa, contains the pterygoid muscles, maxillary artery, mandibular nerve branches, and the pterygoid venous plexus and opens posteriorly into the area around the carotid sheath as shown on the left side. (B) Enlarged view. The soft palate has been divided in the midline, and the leaves reflected laterally. The atlanto-occipital joints and the foramen magnum are located at approximately the level of the hard palate. The anterior arch of C1 and the dens are located behind the oropharynx and the clivus is located behind the nasopharynx and sphenoid sinus. The prominence over the longus capitus and the anterior arch of C1 are seen through the pharyngeal mucosa. (C) The mucosa lining the posterior pharyngeal wall has been reflected to the right, exposing the longus capitus, which attaches to the clivus, and the part of the longus colli that attaches to the anterior arch of C1. The left eustachian tube has been divided. (D) The clivus and anterior arch of C1 have been removed. The dura has been opened to expose the vertebral and basilar arteries. The dens has been preserved. The structures in the right infratemporal fossa and a segment of the right carotid artery and mandible have been removed to expose the right vertebral artery ascending between the C2 and C1 transverse processes. (E) Cross section through the ethmoidal and maxillary sinuses and the nasal cavity in front of the posterior maxillary wall. The posterior wall of the maxillary sinus has been removed to expose the pterygopalatine fossa and ganglia on both sides. The maxillary nerves enter the pterygopalatine fossa by passing through the foramen rotundum. The maxillary arteries enter the pterygopalatine fossa laterally by passing through the pterygomaxillary fissure and give rise to its terminal branches in the pterygopalatine fossa. Another branch enters the greater palatine canal with the greater palatine nerves. (F) Enlarged view of the pterygopalatine fossa. The vidian nerve exits the vidian canal to enter the pterygopalatine ganglion, which receives communicating rami from the maxillary nerve. The sphenopalatine branch passes through the sphenopalatine foramen to enter the lateral nasal cavity (1). Abbreviations: A., artery; Bas., basilar; Cap., capitis; Car., carotid; Cav., cavernous; Comm., communicating; Cond., condyle; Eth., ethmoid; Eust., eustachian; For., foramen; Gang., ganglion; Gr., greater; Inf., inferior; Infraorb., infraorbital; Infratemp., infratemporal; Int., internal; Jug., jugular; Lat., lateral; Lig., ligament; Long., longus; M., muscle; Mandib., mandibular; Mass., masseter; Max., maxillary; Med., medial; Mid., middle; N., nerve; Palat., palatine; Pteryg., pterygoid; Pterygopal., pterygopalatine; Sphen., sphenoid; Sphenopal., sphenopalatine; Sup., superior; Trans., transverse; V., vein; Vert., Vertebral.
Chapter 1: Anatomy of the Cranial Base
A
21
Sphenoid bone
Temp. bone
Occip. bone
Int. Ac. meatus
Jug. For.
Occip. condyle Occip. bone
B
PCA CN III
SCA Bas. A.
CN IV CN V Tent. edge CN VI CN VII, VIII
AICA
CN IX, X
CN XII Vert. A.
P.I.C.A.
CN XI
Figure 12 (A) Superior view of the posterior cranial fossa. The osseous walls of the posterior fossa are formed by the occipital, temporal, and sphenoid bones. The fossa is bounded in front by the dorsum sellae and posterior part of the sphenoid bone and the clival part of the occipital bone; behind by the lower portion of the squamosal part of the occipital bone; and on each side by the petrous and mastoid parts of the temporal bone, and the lateral part of the occipital bone. One small part above the temporal bone is formed by the inferior angle of the parietal bone. (B) Nerves and arteries of the posterior fossa. Only two of the twelve pairs of cranial nerves course entirely outside the posterior fossa. The tentorium, which is attached along the petrous ridges, roofs the posterior fossa. The superior cerebellar artery courses below the oculomotor and trochlear nerves and above the trigeminal nerve; the anteroinferior cerebellar artery courses near the abducens, facial, and vestibulocochlear nerves; and the posteroinferior cerebellar artery courses near the glossopharyngeal, vagus, accessory, and hypoglossal nerves (12). Abbreviations: A., artery; Ac., acoustic; AICA, anteroinferior cerebellar artery; Bas., basilar; CN, cranial nerve; For., foramen; Int., internal; Jug., jugular; Occip., occipital; PCA, posterior cerebral artery; PICA, posteroinferior cerebellar artery; SCA, superior cerebellar artery; Temp., temporal; Tent., tentorial; Vert., vertebral.
22
Martins and Rhoton
A
B Zygomatic proc. Ant. Root Trig. Impress. Squamosal part Groove for Gr. Pet. N. Zygomatic Proc. Post. Root
Petrous part Squamosal part
Petrous part
Mandib. fossa
Arc. Emin. Car. canal
Tymp. Part Styloid proc.
Jug. fossa
Squamosal part Tegmen
Stylomast. For.
Mastoid part
Mastoid part
D
C
Vert. crest
Squamosal part
Facial canal Arc. Emin.
Trans. crest
Sup. Vest. area
Trig. Impress. Sig. sulcus Petrous part Int. Ac. Meatus
Cochlear area Mastoid part
Inf. Vest. area
Vest. Aqueduct
Figure 13 Temporal bone. (A) and (B) Inferior views of temporal bone. (A) The temporal bone has a squamosal part, which forms some of the floor and lateral wall of the middle cranial fossa. It is also the site of the mandibular fossa in which the mandibular condyle sits. The tympanic part forms the anterior, lower, and part of the posterior wall of the external canal, part of the wall of the tympanic cavity, the osseous portion of the eustachian tube, and the posterior wall of the mandibular fossa. The mastoid portion contains the mastoid air cells and mastoid antrum. The petrous part is the site of the auditory and vestibular labyrinth, the carotid canal, the internal acoustic meatus, and the facial canal. The petrous part also forms the anterior wall and the dome of the jugular fossa. The styloid part projects downward and serves as the site of attachment of three muscles. (B) Superior view. The medial part of the upper surface is the site of the trigeminal impression in which Meckel’s cave sits. Further laterally is the prominence of the arcuate eminence overlying the superior semicircular canal. Anterior and lateral to the arcuate eminences is the tegmen, a thin plate of bone overlying the mastoid antrum and epitympanic area. The temporal bone articulates anteriorly with the sphenoid bone, above with the parietal bone, and posteriorly with the occipital bone. The zygomatic process of the squamosal part has an anterior and a posterior root between which, on the lower surface, is located the mandibular condyle. (C) Posterior view of a right temporal bone. The sigmoid sulcus descends along the posterior surface of the mastoid portion. The internal acoustic meatus enters the central portion of the petrous part of the bone. The trigeminal impression and arcuate eminence are located on the upper surface of the petrous part. The vestibular aqueduct connects the vestibule in the petrous part with the endolymphatic sac, which sits on the posterior petrous surface inferolateral to the internal acoustic meatus. (D) Enlarged view of the right internal acoustic meatus. The transverse crest divides the meatal fundus into superior and inferior parts. The anterior part above the transverse crest is the site of the facial canal and the posterior part is the site of the superior vestibular area. Below the transverse crest the cochlear area is anterior and the inferior vestibular area is posterior. The vertical crest, also called “Bill’s Bar,” separates the facial and superior vestibular areas (7). Abbreviations: Ac., acoustic; Ant., anterior; Arc., arcuate; Car., carotid; Emin., eminence; For., foramen; Gr., greater; Impress., impression; Inf., inferior; Int., internal; Jug., jugular; Mandib., mandibular; N., nerve; Pet, petrosal; Post, posterior; Proc., process; Sig., sigmoid; Stylomast., stylomastoid; Sup., superior; Trans., transverse; Trig., trigeminal; Tymp., tympanic.; Vert., vertebral; Vest., vestibular.
Chapter 1: Anatomy of the Cranial Base
23
A Basal Part (Clivus)
Car. canal
Styloid Proc.
B Clivus
Jug. For. Stylomast. For. Occip. Cond.
Condylar part
Petrocliv. Fiss. Jug. tubercle
Jug. For. Occip. Cond.
Digast. groove Sig. sulcus
Occip. A. groove
Vermian fossa Ext. Occip. crest Int. Occip. crest Sup. nuchal line Squamosal part Inion
Sup. sag. Sinus sulcus
Trans. sinus sulcus
D
C
For. ovale For. lacerum Clivus
Petrocliv. Fiss. Styloid Proc.
Pharyng. tubercle Jug. For.
Styloid Proc.
Stylomast. For.
Occip. Cond.
Car. canal
Clivus
Jug. For.
Condylar part
21 mm
Occip. Cond. 8 mm
Occip. bone Jug. Proc.
Squamosal Part
E
F For. Ovale Int. Ac. meatus
Petrocliv. Fiss.
Clivus
Coch. aqueduct Pet. part
Clivus
Vest. aqueduct
Petrocliv. Fiss. Car. canal
Coch. aqueduct Intrajug. Proc. Sig. Part
Hypogl. canal Occip. Cond.
Pet. part Sig. part Occip. bone Jug. Proc.
Figure 14 Occipital bone, foramen magnum, and jugular foramen. (A–D) Occipital bone and foramen magnum. (A) Inferior view. (B) Superior view. (C, D) Anteroinferior views. (A–C) The occipital bone surrounds the oval shaped foramen magnum, which is wider posteriorly than anteriorly. The narrower anterior part sits about the odontoid process and is encroached on from laterally by the occipital condyles. The wider posterior part transmits the medulla. The occipital bone is divided into a squamosal part located above and behind the foramen magnum; a basal (clival) part situated in front of the foramen magnum; and paired condylar parts located lateral to the foramen magnum. The basilar part of the occipital bone, which is also referred to as the clivus, is a thick quadrangular plate of bone, concave from side to side, that extends forward and upward to join the sphenoid bone just below the dorsum sellae. The clivus is separated on each side from the petrous part of the temporal bone by the petroclival fissure, which ends posteriorly at the jugular foramen. The condylar parts of the occipital bone, on which the occipital condyles are located, are situated lateral to the foramen magnum on the external surface. The hypoglossal canal is situated above the condyle. The jugular process of the occipital bone extends laterally from the posterior half of the condyle and articulates with the jugular surface of the temporal bone. The sulcus of the sigmoid sinus crosses the superior surface of the jugular process. The jugular foramen is bordered posteriorly by the jugular process of the occipital bone and anteriorly by the jugular fossa of the petrous temporal bone. (Continued).
24
Martins and Rhoton
ring, inside the collar of dura, and around the clinoid segment to the level of the upper ring. The meningohypophyseal trunk, with its tentorial, inferior hypophyseal, and dorsal meningeal branches, and the inferolateral trunk, also called the artery of the inferior cavernous sinus, arise from the intracavernous carotid artery. The proximal abducens nerve passes through Dorello’s canal, located below the petrosphenoid ligament, and receives sympathetic branches from the internal carotid nerve, which pass to the ophthalmic nerve to enter the orbit. The main venous afferents to the cavernous sinus are the superior and inferior ophthalmic veins and the sphenoparietal sinus (Figs. 4 and 7). Several venous compartments, named according to their relationship to the cavernous carotid artery, empty mainly in the basilar and superior and inferior petrosal sinuses, or, by way of the foramina in the middle fossa floor, into the pterygoid venous plexus (4). The sella houses the pituitary gland and is partially closed above by the diaphragma sellae. Anterolateral to the diaphragm, the carotid cave, a dural depression at the level of the distal dural ring, extends downward medial to the initial intradural segment of the internal carotid artery. The tensor tympani muscle and eustachian tube cross medial to the foramen spinosum, below the floor of the middle fossa, and anterior to the horizontal segment of the petrous carotid (Fig. 5). The greater petrosal nerve crosses the area above and parallel to the petrous carotid artery, laterally joins the geniculate ganglion, and medially joins the deep petrosal branch of the carotid sympathetic nerves to form the vidian nerve, which enters the pterygopalatine ganglion (Figs. 2, 5, and 6). The lesser petrosal nerve runs anterior to the greater petrosal nerve and exits the cranium, passing through the foramen spinosum to join the otic ganglion. The cochlea is situated below the floor of the middle cranial fossa, at the apex of the angle between the greater petrosal and labyrinthine segment of the facial nerve.
Exocranial Surface The exocranial surface of the middle cranial base is also divided into central and lateral parts (Figs. 2, 3, 6, and 9) (1). The central part encompasses the sphenoid body and the upper part of the basal (clival) part of the occipital bone and corresponds to the sphenoid sinus and the nasopharynx. The lateral part is formed by the greater sphenoid wing; the petrous, tympanic, and squamous parts of the temporal bone; the styloid process; and the zygomatic, palatine, and maxillary bones. The medial and lateral parts are separated by a parasagittal plane passing through the medial pterygoid plate. The foramen lacerum is located at the union of the sphenoid, occipital, and petrous bones and is enclosed on its lower side by fibrocartilaginous tissue to form the inferior wall of the carotid canal. Structures transversing the lateral part include the carotid artery in the carotid canal, the glossopharyngeal, vagus, and accessory nerves in the jugular foramen, the third trigeminal division in the foramen ovale, the middle meningeal artery in the foramen spinosum, and the facial nerve in the facial canal. The pterygomaxillary fissure is the lateral opening of the pterygopalatine fossa into the infratemporal fossa. The glenoid fossa harbors the mandibular condyle. The roof of the fossa is divided into anterior and posterior parts by the squamotympanic fissure, along which the chorda tympani passes. The area below the middle cranial base includes the infratemporal fossa, parapharyngeal space, infrapetrosal space, and pterygopalatine fossa (Figs. 6, 9, 10, and 11). The boundaries of the infratemporal fossa are the middle pterygoid muscle and the pterygoid process medially, the mandible laterally, the posterior wall of the maxillary sinus anteriorly, the greater wing of the sphenoid superiorly, and the medial pterygoid muscle joining the mandible and the pterygoid fascia posteriorly. The fossa opens into the neck below. The infratemporal fossa contains the branches of mandibular nerve, the maxillary artery, and the pterygoid muscles and venous plexus. The mandibular nerve, after exiting the
←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− Figure 14 (Continued) The jugular tubercle lies on the internal surface above the hypoglossal canal. The squamous part is internally concave. The convex external surface of the squamosal part has several prominences. The largest prominence, the external occipital protuberance (inion), is situated at the central part of the external surface. The superior nuchal line radiates laterally from the protuberance. A vertical ridge, the external occipital crest, descends from the external occipital protuberance to the midpoint of the posterior margin of the foramen magnum. The internal surface of the squamous part has a prominence, the internal occipital protuberance near its center. The internal occipital crest bifurcates above the foramen magnum to form a V-shaped ridge between the limbs of which is the vermian fossa. (D) Inferior view of the occipital condyles and foramen magnum. The occipital condyles are located along the lateral margin of the anterior half of the foramen magnum. Their articular surfaces are convex, face downward and laterally, and articulate with the superior facet of C1. A probe inserted through the hypoglossal canal passes forward approximately 45 degrees from the midsagittal plane in an anterolateral direction. The hypoglossal canal is located above the middle third of the occipital condyle and is directed from posterior to anterior and from medial to lateral. The intracranial end of the hypoglossal canal is located approximately 5 mm above the junction of the posterior and middle third of the occipital condyle and approximately 8 mm from the posterior edge of the condyle. The extracranial end of the canal is located approximately 5 mm above the junction of the anterior and middle third of the condyle. The average length of the longest axis of the condyle is 21 mm. The large arrow shows the direction of the transcondylar approach and the cross-hatched area shows the portion of the occipital condyle that can be removed without exposing the hypoglossal nerve in the hypoglossal canal. The stylomastoid foramen is situated lateral to the jugular foramen. The styloid process is located anterior and slightly medial to the stylomastoid foramen. (E, F) Jugular foramen. (E) Posterosuperior view. The jugular foramen is located between the temporal and occipital bones. The sigmoid groove descends along the mastoid and crosses the occipitomastoid suture where it turns forward on the upper surface of the jugular process of the occipital bone and enters the foramen by passing under the posterior part of the petrous temporal bone. The foramen has a larger lateral sigmoid part through which the sigmoid sinus empties and a smaller anteromedial petrosal part through which the inferior petrosal sinus empties. The two parts are separated by the intrajugular processes of the occipital and temporal bones. The glossopharyngeal, vagus, and accessory nerves pass through the intrajugular portion of the foramen located between the sigmoid and petrosal parts. The foramen is asymmetric from side to side, with the right side often being larger as shown. The cochlear aqueduct opens just above the anterior edge of the petrosal part. The vestibular aqueduct opens into the endolymphatic sac, which sits on the back of the temporal bone superolateral to the sigmoid part of the jugular foramen. (F) Anteroinferior view. The roof over the jugular foramen, formed by the jugular fossa of the temporal bone, is shaped to accommodate the jugular bulb. The posterior margin of the foramen is formed by the jugular process of the occipital bone, which connects the basal (clival) part of the occipital bone to the squamosal part. The petroclival fissure intersects the anteromedial margin of the petrosal part of the foramen. The entrance into the carotid canal is located directly in front of the medial half of the jugular foramen. (A–C) (13), (D) (14), (E–G) (18). Abbreviations: A., artery; Ac., acoustic; Car., carotid; Coch., Cochlear; Cond., condyle; Digast., digastric; Ext., external; Fiss., fissure; For., foramen; Hypogl., hypoglossal; Int., internal; Intrajug., intrajugular; Jug., jugular; Occip., occipital; Petrocliv., petroclival; Pet., petrous; Pharyng., pharyngeal; Proc., process; Sag., sagittal; Sig., sigmoid; Stylomast., stylomastoid; Sup., superior; Trans., transverse; Vest., vestibular.
Chapter 1: Anatomy of the Cranial Base
A
25
B Sup. Pet. V. Subarc. A. CN VII
CN V
CN VIII
Sup. Vest. N. Nerv. Intermed. Inf. Vest. Coch. N.
Labyr. A. A.I.C.A.
Flocc.
Flocc. CN VI CN IX CN IX PICA
PICA CN X
CN X-XI
Chor. Plex.
CN XI
D
C
CN V CN V
CN VII Sup. Vest. N. Nerv. Intermed.
Coch. N.
Sup. Vest. N.
Inf. Vest. N. AICA
Subarc. A. CN VII Inf. Vest. N. Coch. N. Labyr. A.
Flocc.
Inf. Vest. N. Flocc.
CN VI
CN IX
CN VI
CN IX
Chor. Plex. PICA
AICA
Chor. Plex. PICA
Figure 15 Retrosigmoid exposure of the nerves in the right cerebellopontine angle. (A) The vestibulocochlear nerve enters the internal acoustic meatus with a labyrinthine branch of the AICA. The PICA courses around the glossopharyngeal, vagus, and accessory nerves. The abducens nerve ascends in front of the pons. The subarcuate branch of the AICA enters the subarcuate fossa superolateral to the porus of the meatus. Choroid plexus protrudes into the cerebellopontine angle behind the glossopharyngeal and vagus nerves. (B) The posterior wall of the internal acoustic meatus has been removed. The cleavage plane between the upper bundle, formed by the superior vestibular nerve, and the lower bundle, formed by the inferior vestibular and cochlear nerves, was begun laterally where the nerves normally separate near the meatal fundus and extended medially. The nervus intermedius arises on the anterior surface of the vestibulocochlear nerve, has a free segment in the cistern and/or meatus, and joins the facial nerve distally. The facial nerve is located anterior to the superior vestibular nerve and the cochlear nerve is anterior to the inferior vestibular nerve. (C) The cleavage plane between the cochlear and inferior vestibular nerves, which is well developed in the lateral end of the internal acoustic meatus, has been extended medially. Within the cerebellopontine angle, the superior vestibular nerve is posterior and superior, the facial nerve anterior and superior, the inferior vestibular nerve posterior and inferior, and the cochlear nerve anterior and inferior. (D) The superior and inferior vestibular nerves have been divided to expose the facial and cochlear nerves. A labyrinthine branch of the AICA enters the internal meatus (15). Abbreviations: A., artery; AICA, anteroinferior cerebellar artery; Chor. Plex., choroid plexus; CN, cranial nerve; Coch., cochlear; Flocc., flocculus; Inf., inferior; Intermed., intermedius; Labyr., labyrinth; N., nerve; Nerv., nervus; Pet., petrosal; PICA, posteroinferior cerebellar artery; Subarc., subarcuate; Sup., superior; V., vein; Vest., vestibular.
26
Martins and Rhoton
B
A Pon. Mes. Sulcus
CN VI Pon. Med. Sulcus CN IX, X, XI Olive Medulla
CN VII CN VII, VIII CN VIII Chor. Plex. Pyramids AICA Pet. Surface CN IX, X CN XII CN XII CN XI PICA Vert. A. PCA CN IV SCA
PCA CN III SCA
D
AICA
PICA
CN XII
Ant. Hem. V.
Ant. Sp. A.
PICA Vert. A.
Vert. A.
CN V CN VI AICA
PICA
Ped. V.
CN VI Trans. Pon. V. AICA Sup. Pet. V. CN IX -XI
CN VII, VIII
CN III SCA
CN V
Pons
C
CN V
CN IV
Midbrain
Med. Ant. Pon. Mes. V. V. Cer. Pon. Fiss.
Ant. Hem. V. V. Pon. Med. sulcus Trans. Med. V.
Med. Ant. Med. V.
Figure 16 Brain stem, anterior cerebellar surface, and posterior skull base. (A) The petrosal (anterior) surface of the cerebellum, called the petrosal surface, and front of the brain stem face the endocranial surface of the posterior fossa. The fourth ventricle is positioned behind the pons and medulla. The midbrain and pons are separated by the pontomesencephalic sulcus and the pons and medulla by the pontomedullary sulcus. The trigeminal nerves arise from the midpons. The abducens nerve arises in the medial part of the pontomedullary sulcus, rostral to the medullary pyramids. The facial and vestibulocochlear nerves arise at the lateral end of the pontomedullary sulcus immediately rostral to the foramen of Luschka. The hypoglossal nerves arise anterior to the olives and the glossopharyngeal, vagus, and accessory nerves arise posterior to the olives. Choroid plexus protrudes from the foramen of Luschka behind the glossopharyngeal and vagus nerves. (B) Anterior view of the brain stem with the arteries preserved. (C) Posterior view of the skull base with the cranial nerves and arteries preserved. (B, C) The SCA arises at the midbrain level and encircles the brain stem near the pontomesencephalic junction. The SCA courses below the oculomotor and trochlear nerves and above the trigeminal nerve. The SCA loops down closer to the trigeminal nerve in C than in B. The AICA arises at the pontine level and courses by the abducens, facial, and vestibulocochlear nerves. In B, both AICAs pass below the abducens nerves. In C, the left abducens nerve passes in front of the AICA and the right abducens nerve passes behind the AICA. The PICAs arise from the vertebral artery at the medullary level and course in relation to the glossopharyngeal, vagus, accessory, and hypoglossal nerves. The origin of the SCAs is quite symmetrical from side to side. There is slight asymmetry in the level of origin of the AICAs and marked asymmetry in the level of the origin of the PICAs, especially in B. (D) The veins on the anterior surface of the pons and medulla and the petrosal cerebellar surface drain predominantly into the superior petrosal veins which empty into the superior petrosal sinuses. The median anterior pontomesencephalic and median anterior medullary veins ascend on the front of the brain stem. The transverse pontine and transverse medullary veins run transversely across the pons and medulla surfaces. The anterior hemispheric veins drain the petrosal cerebellar surface and commonly empty into the vein of the cerebellopontine fissure, which ascends to join the superior petrosal veins. The vein of the pontomedullary sulcus passes across the pontomedullary junction. The peduncular vein crosses the cerebral peduncle. Abbreviations: A., artery; AICA, anteroinferior cerebellar artery; Ant., anterior; Cer. Pon., cerebellopontine; Chor., choroid; CN, cranial nerve; Fiss., fissure; Hem., hemispheric; Med., medial, medullary; Pet., petrosal; PCA, posterior cerebral artery; Ped., peduncular; PICA, posteroinferior cerebellar artery; Plex., plexus; Pon., pontine; Pon. Med., pontomedullary; Pon. Mes., pontomesencephalic; SCA, superior cerebellar artery; Sulc., sulcus; Sp., spinal; Sup., superior; Trans., transverse; V., vein; Vert., vertebral.
foramen ovale, lies anterolateral to the otic ganglion and divides immediately into its terminal branches: the pterygoid, buccal, masseteric, and temporal branches along the superior wall of the fossa; the inferior alveolar and the lingual branches, after being joined by the chorda tympani, descend between both pterygoid muscles; and the auriculotemporal branch with the maxillary artery course between the mandible and the sphenomandibular ligament. The auriculotemporal nerve carries the parasympathetic innervation of
the parotid gland, which travels through the tympanic branch of the glossopharyngeal nerve, which in turn forms the lesser petrosal nerve, to reach the otic ganglion before joining the auriculotemporal nerve. The maxillary artery, which arises as a terminal branch of the external carotid artery with the superficial temporal artery, is divided into three segments. The first, or mandibular segment, passes between the sphenomandibular ligament and the mandibular neck and gives rise to the deep auricular, anterior tympanic, middle meningeal,
Chapter 1: Anatomy of the Cranial Base
A
27
B Post. Digast. M. Rec. Cap. Post. Maj. M.
Sup. Obl. M. Suboccip. triangle
Trans. Proc. C1
Inf. Obl. M.
X
Lev. Scap. M.
C
C
CN IX, X
D
th P.I.C.A. 4 Vent.
PICA
CN XI CN XI
Occip. Cond. Atl. Occip. Joint
CN XII C1 Cond. CN XII Vert. A. Dural cuff
Dent. Lig.
C1 Cond. Vert. A. C1 Trans. Proc. C1
Hypogl. canal
Vert. A. Vert. A.
Figure 17 (A–D) Far-lateral and transcondylar approach. (A) A suboccipital scalp flap is commonly selected for the far-lateral exposure. The medial limb extends downward in the midline so that a wide upper cervical laminectomy can be completed if needed. The lateral limb extends below the level of the C1 transverse process (×), which can be palpated between the mastoid tip and the angle of the jaw, to access the vertebral artery as it ascends through the C1 transverse process. The muscles superficial to the suboccipital triangle can be reflected from the suboccipital area in a single layer with the scalp flap, leaving a cuff of suboccipital muscle and fascia attached along the superior nuchal line to aid in closure. (B) The scalp and muscles are reflected in a single layer to expose the suboccipital triangle in the depths of which the vertebral artery courses behind the atlanto-occipital joint and across the posterior arch of C1. The triangle is located between the superior and inferior oblique and the rectus capitis posterior major. (C) A suboccipital craniectomy has been completed, the posterior arch of C1 has been removed, the posterior root of the transverse foramen of the C1 has been removed, the area above the occipital condyle has been drilled to expose the hypoglossal canal, and the dura has been opened. The dural incision completely encircles the vertebral artery, leaving a narrow dural cuff on the artery, so that the artery can be mobilized. The drilling in the supracondylar area can be extended extradurally to the level of the jugular tubercle to increase access to the front of the brain stem. (D) Comparison of the exposure with the far-lateral and transcondylar approaches. On the right side, the far-lateral exposure has been extended to the posterior margins of the atlantal and occipital condyles and the atlanto-occipital joint. The prominence of the condyles limits the exposure along the anterolateral margin of the foramen magnum. On the left side, a transcondylar exposure has been completed by removing the posterior part of the condyles. The dura can be reflected further laterally with the transcondylar approach than with the far-lateral approach. The condylar drilling provides an increased angle of view of the clivus and front of the brain stem. The dentate ligament and accessory nerve ascend from the region of the foramen magnum (18). Abbreviations: A., artery; Atl., atlanto; Cap., capitis; CN, cranial nerve; Cond., condyle; Dent., dentate; Digast., digastric; Hypogl., hypoglossal; Inf., inferior; Lev., levator; Lig., ligament; M., muscle; Maj., major; Obl., oblique; Occip., occipital; PICA, posteroinferior cerebellar artery; Post., posterior; Proc., process; Rec., rectus; Scap., scapula; Suboccip., suboccipital; Sup., superior; Trans., transverse; Vent., ventricle; Vert., vertebral.
accessory middle meningeal (enters through the foramen ovale), and the inferior alveolar artery. The second, or pterygoid segment, courses through the middle of the infratemporal fossa and gives rise to the posterosuperior alveolar, infraorbital, masseteric, pterygoid, temporal, and buccal branches. The third, or pterygopalatine segment, courses in
the fossa of the same name. The pterygoid venous plexus connects through the middle fossa foramina and inferior orbital fissure with the cavernous sinus and empties into the retromandibular and facial veins. The pterygopalatine fossa is located between the maxillary sinus in the front, the pterygoid process behind, the
28
Martins and Rhoton
Ant. tubercle Ant. arch
A
B
Ant.tubercle
Ant. arch
Trans. proc.
Sup. Art. Facet Trans. Proc.
Inf. Art. Facet
Lat. mass. Trans. For. Trans. For. Post. arch
Groove for Vert. A.
Post. arch
Post. tubercle
C
Post. tubercle
D
Sup. Art. Facet
Sup. Art. Facet
Trans. Proc. Groove for Vert. A. Trans. Proc.
Post. arch
Ant. arch
Ant. tubercle
Inf. Art. Facet
E
Odontoid proc.
Post. tubercle
Inf. Art. Facet
F Odontoid proc.
Art. facet Sup. Art. Facet
Spinous proc.
Sup. Art. Facet
Sup. Art. Facet Trans. For.
Trans. proc.
Lamina
Trans. for.
Trans. proc. Body
Body Trans. proc.
Inf. Art. Facet Inf. Art. Facet
Inf. Art. Facet
G
Odontoid proc. Sup. Art. Facet Trans. proc.
Sup. Art. Facet Trans. proc.
Body
H Trans. for Trans. ror.
Body
Trans. proc.
Trans. for. Inf. Art. Facet
Pedicle
Inf. Art. Facet
Inf. Art. Facet Inf. Art. Facet Lamina
Lamina
Spinous proc. Spinous proc.
Figure 18 Atlas and Axis. (A–D) The atlas. (A) Superior view. (B) Inferior view. (C) Anterior view. (D) Posterior view. The atlas consists of two thick lateral masses situated at the anteromedial part of the ring, and which are connected in front by a short anterior arch and posteriorly by a longer curved posterior arch. The anterior and posterior tubercles are at the anterior and posterior midline. The superior articular facet is an oval, concave facet that faces upward and medially to articulate with the occipital condyle. The inferior articular facet is a circular, flat, or slightly concave facet that faces downward, medially, and slightly backward and articulates with the superior articular facet of the axis. The medial aspect of each lateral mass has a small tubercle for the attachment of the transverse ligament of the atlas. The transverse process projects from the lateral masses. The transverse foramina transmit the vertebral arteries. The upper surface of the posterior arch adjacent to the lateral masses has paired grooves in which the vertebral arteries course. (E–H) The axis. (E) Anterior view. (F) Lateral view. (G) Superior view. (H) Inferior view. The axis is distinguished by the odontoid process (dens). On the front of the dens is an articular facet that forms a joint with the facet on the back of the anterior arch of the atlas. The dens is grooved at the base of its posterior surface where the transverse ligament of the atlas passes. The oval superior articular facets articulate with the inferior facets of the atlas. The superior facets are anterior to the inferior facets. The pedicles and laminae are thicker than on the other cervical vertebra and the lamina fuses behind to form a large spinous process. The transverse foramina are directed superolaterally, thus permitting the lateral deviation of the vertebral arteries as they pass up to the more widely separated transverse foramina in the atlas. The inferior articular facets face downward and forward (13). Abbreviations: A., artery; Ant., anterior; Art., articular; For., foramen; Inf., inferior; Lat., lateral; Mass., masses; Post., posterior; Proc., process; Sup., superior; Trans., transverse; Vert., vertebral.
Chapter 1: Anatomy of the Cranial Base
29
Nasal bone B
A Zygoma
Nasal bone
Maxilla Med. Canthal Lig.
Front. bone
Nasolac. duct Lac. bone Periorbita
D
C Maxilla
Nasal cavity Med. Canthal Lig.
Nasolac. duct
Sup. concha
Lac. bone Nasal bone Eth. Perp. plate Ant. Eth. A. Eth. sinus Crib. plate
Frontonasal suture
Periorbita
Dura
E
F Sphenopal. A. Clivus
Vomer Ant. Eth. A.
Septa Car. A.
Sphen. sinus
Sphen. ostia Sphen. sinus Crib. plate Crib. plate
Figure 19 (A–F). Relationships in the transbasal and extended frontal approaches. (A) The inset shows the bicoronal scalp incision. A large bifrontal craniotomy and a fronto-orbitozygomatic osteotomy have been completed. The osteotomized segment may extend through the nasal bone and from one to the other lateral orbital rims, as shown. However, for most lesions, a more limited bone flap and osteotomy will suffice and can be tailored as needed to deal with the involvement of the cranial base, nasal cavity, paranasal sinuses, or orbit. For an orbital lesion, an orbitofrontal craniotomy, elevating only the superior orbital rim (yellow arrows) and orbital roof, is all that is needed. For a cavernous sinus or unilateral lesions of the anterior or middle fossa, an orbitozygomatic osteotomy will usually suffice (blue arrow). For a clival lesion, a more limited bifrontal approach (red arrow) will suffice. (B) The periorbita has been separated from the walls of the orbit in preparation for the osteotomies. Division of the medial canthal ligament is not necessary for most lesions, but may be required for lesions extending into the lower nasal cavity or orbit. The ligaments should be re-approximated at the end of the operation. (C) The right medial canthal ligament has been divided and the orbital contents retracted laterally to expose the nasolacrimal duct and the anterior ethmoidal branch of the ophthalmic artery at the anterior ethmoidal foramen. (D) The osteotomies have been completed and the frontal dura elevated. The dura remains attached at the cribriform plate. The upper part of both orbits are exposed. (E) An osteotomy around the cribriform plate leaves it attached to the dura and olfactory bulbs, a maneuver that has been attempted in order to preserve olfaction but has been uncommonly successful. The anterior face of the sphenoid sinus and both sphenoid ostia are exposed between the orbits. (F) The sphenoid sinus has been opened to expose the septa within the sinus. The sphenopalatine arteries cross the anterior face of the sphenoid. (Continued).
30
Martins and Rhoton
G
H
Vomer
Vomer Clival dura
Car. A. Cav. Seg.
Clivus
Car. A. Optic canal
Bas. sinus Pit. gland
Pit. gland Optic canal
I
J Med. Pteryg. plate Olf. Tr.
Eust. tube Pteryg. Proc. Max. A. Pterygopal. Gang. Infraorb. N. Sphen. Sinus
CN II Chiasm MCA
MCA
V2 ACA
Car. A. Figure 19 (Continued) (G) The septa within the sphenoid sinus, the sellar floor, and the lateral sinus wall have been removed to expose the intracavernous carotid, pituitary gland, and optic canals. (H) The clivus has been opened to expose the dura facing the brain stem. The basilar sinus, which interconnects the posterior parts of the cavernous sinus, is situated between the layers of dura on the upper clivus. (I) The exposure has been extended laterally by opening the medial and posterior wall of the maxillary sinus to expose the branches of the maxillary nerve and artery in the pterygopalatine fossa, located behind the posterior maxillary wall. The posterior wall of the pterygopalatine fossa is formed by the pterygoid process. The maxillary nerve enters the pterygopalatine fossa where it gives rise to the infraorbital nerve, which courses along the floor of the orbit and to the palatine nerves, which descend to the palatal area. The eustachian tube opens into the nasopharynx by passing along the posterior edge of the medial pterygoid plate. The lateral wing of the sphenoid sinus extends laterally below the maxillary nerve. (J) The frontal dura has been opened and the frontal lobes elevated to expose the olfactory and optic nerves and the internal carotid and anterior and middle cerebral arteries (1). Abbreviations: A., artery; ACA, anterior cerebral artery; Ant., anterior; Bas., basilar; Car., carotid; Cav., cavernous; CN cranial nerve; Crib., cribriform; Eth., ethmoid, ethmoidal; Eust., eustachian; Front., frontal; Gang., ganglion; Infraorb., infraorbital; Lac., lacrimal; Lig., ligament; Max., maxillary; Med., medial; MCA, middle cerebral artery; N., nerve; Nasolac., nasolacrimal; Olf., olfactory; Perp., perpendicular; Pit., pituitary; Proc., process; Pteryg., pterygoid; Pterygopal., pterygopalatine; Seg., segment; Sphen., sphenoid; Sphenopal., sphenopalatine; Sup., superior; Tr., Tract.
palatine bone medially and the body of the sphenoid bone above (Figs. 3, 6, 10, and 11). The fossa opens laterally through the pterygomaxillary fissure into the infratemporal fossa and medially through the sphenopalatine foramen to the nasal cavity. Both the foramen rotundum for the maxillary nerve and the pterygoid canal for the vidian nerve open through the posterior wall of the fossa formed by the sphenoid pterygoid process. The palatovaginal canal carrying the pharyngeal nerve and artery and the greater and lesser palatine canals conveying the greater and lesser palatine arteries open into the pterygopalatine fossa. The inferior orbital fissure, across which the orbital muscle stretches, lies in front of the pterygopalatine fossa. This fossa contains branches of the maxillary nerve, vidian nerve, the pterygopalatine ganglion, and the pterygopalatine segment of the maxillary artery. The
maxillary nerve passes through the foramen rotundum to enter the fossa and, after giving communicating rami to the pterygopalatine ganglion, divides into the posterosuperior alveolar, infraorbital, and zygomatic nerves. The zygomatic nerve, in addition to its sensory fibers, carries the parasympathetic fibers from the pterygopalatine ganglion to the lacrimal gland. The vidian (nerve of the pterygoid canal) ends in the pterygopalatine ganglion, which sends rami to the maxillary nerve and gives rise to the greater and lesser palatine, pharyngeal nerves, and nasal branches. The third part of the maxillary artery enters the fossa and divides into its terminal lesser and greater palatine, sphenopalatine, vidian, and pharyngeal branches. The parapharyngeal space lies between the structures in the pharynx wall medially, the medial pterygoid muscle
31
Chapter 1: Anatomy of the Cranial Base
A
C
B
Hard Palate Clivus
Clivus
Soft palate reflected Long. Cap. M.
Soft palate Uvula For. magnum
Ant. Arch C1
Ant. Arch C1
Dens Long. Colli M.
D
E Nasal septum
Osteotomy
Vert. A.
Max. sinus
E
Maxilla Max. sinus Gr. Palat. A. & N.
Nasal septum Max. sinus
Nasal floor
Figure 20 (A) Anterior view through the open mouth. The soft palate, which extends backward from the hard palate, will block the view of the upper clivus. An incision has been outlined in the midline of the soft palate. (B) The soft palate has been divided to expose the mucosa lining the lower clivus. (C) The pharyngeal mucosa has been opened in the midline and the left longus capitus and longus coli have been reflected laterally. (D) The transverse maxillary (Le Fort I) osteotomy extends through the maxillary sinus above the apex of the teeth and below the infraorbital canals. (E) The lower maxilla has been displaced downward. A clival window and vertebral arteries are seen through the exposure (1). Abbreviations: A., artery; Ant., anterior; Cap., capitis; For., foramen; Gr., greater; Long., longus; Max., maxillary; M., muscle; N., nerve; Palat., palatine; Vert., vertebral.
and the parotid fascia laterally, and the styloid fascia investing the styloglossus, stylopharyngeal, and the stylohyoid muscles posteriorly (Fig. 6). In its upper medial wall, the eustachian tube, covered below by the tensor and levator veli palatine muscles, runs from the tympanic cavity to the pharyngeal wall. This is predominantly a fat-filled space, but also contains pharyngeal branches of the ascending pharyngeal and facial arteries and branches from the glossopharyngeal nerve. The last of the four spaces below the middle fossa is the infrapetrosal space, also referred to as the poststyloid part of the parapharyngeal space. It is located behind the styloid fascia, below the petrous bone, and medial to the mastoid process (Figs. 2, 6, and 9). Among the foramina in the area connecting the intra- and extracranial spaces is the jugular foramen containing the jugular bulb and lower end of the inferior petrosal sinus. It also contains branches of the ascending pharyngeal artery, the glossopharyngeal, vagus, and accessory nerves, and the opening of the carotid canal through which the carotid artery and the carotid sympathetic
nerves pass. Two tiny foramina located between the jugular foramen and carotid canal carry the tympanic branch of the glossopharyngeal nerve and the auricular branch of the vagus nerve. The stylomastoid foramen, conveying the facial nerve and the stylomastoid artery, opens between the mastoid tip and the styloid processes. The main fissure in the area is the petroclival fissure on the upper and lower side of which courses the inferior petrosal sinus and the inferior petroclival vein, respectively. The main nerves of the area are the glossopharyngeal nerve coursing below the styloglossus muscle, the vagus nerve descending between the internal carotid artery and the jugular vein, and the accessory nerve passing lateral to the jugular vein on its way to the sternocleidomastoid muscle. The facial nerve runs to the parotid gland where it divides into cervicofacial and temporofacial trunks. The hypoglossal nerve, after exiting the hypoglossal canal, descends between the carotid artery and the jugular vein, turning anteriorly across the lateral wall of the artery below the level of the digastric muscle. The main arteries in the area are the internal carotid artery with its cervical
32
Martins and Rhoton
A Front. bone
Orb. Oculi M.
Nasal bone Eth. bone
C
B
Lac. bone Med. canthal Lig.
Lac. Sac
Lac. Canalic.
Maxilla Max. sinus
Zygoma
Nasolac. duct Nasal cavity
D
Lac. sac Med. canthal Lig.
Lac. Canalic.
Max. sinus
F
E
Pteryg. Gang.
V2 Pterygopal. fossa Nasolac. duct Nasal cavity Max. A.
Pons
Bas. A.
Pteryg. Proc.
Inf. concha Inf. meatus
Pit. gland
Long. Cap. M.
Gr. Palat. N.
Figure 21 Relationships of the medial orbit. (A) The medial part of the orbital rim is formed by the frontal bone and maxilla. The anterior part of the nasolacrimal canal is formed by the maxilla and the posterior part by the lacrimal bone, which joins the ethmoid bone posteriorly and the frontal bone above. (B) The medial part of the orbicularis oculi muscle has been exposed. The anterior band of the medial canthal ligament, which crosses in front of the lacrimal sac, is attached to the frontal process of the maxilla medially and to the superior and inferior tarsi laterally. (C) The medial part of the orbicularis oculi muscle and some of the maxilla have been removed to expose the lacrimal sac, nasolacrimal duct, and a small part of the nasal cavity and maxillary sinus. (D) The anterior band of the medial canthal ligament has been reflected laterally to expose the superior and inferior lacrimal canaliculi joining the lacrimal sac. Additional maxilla has been removed to expose the nasal cavity and inferior turbinate medially and the maxillary sinus laterally. The nasolacrimal duct opens into the inferior nasal meatus. (E) Some of the posterior and medial wall of the maxillary sinus has been removed to expose the pterygopalatine fossa, which contains the maxillary nerve and artery and their branches and the pterygopalatine ganglion. (F) The approach has been directed through the nasal cavity medial to the pterygopalatine ganglion and fossa to the clivus, which has been opened to expose the basilar artery (1). Abbreviations: A., artery; Bas., basilar; Canalic., canaliculi; Cap., capitis; Eth., ethmoid; Front., frontal; Gang., ganglion; Gr., greater; Inf., inferior; Lac., lacrimal; Lig., ligament; Long., longus; M., muscle; Max., maxillary; Med., medial; N., nerve; Nasolac., nasolacrimal; Orb., orbital; Palat., palatine; Pit., pituitary; Proc., process; Pteryg., pterygoid; Pterygopal., pterygopalatine.
and petrous segments. The branches of the petrous segment are the caroticotympanic and vidian arteries. The ascending pharyngeal artery ascends medial to the carotid artery, giving meningeal branches which pass through the hypoglossal canal and jugular foramen as well as pharyngeal branches. The occipital artery passes posteriorly on the medial side of the posterior belly of the digastric muscle. The veins in the area are the internal jugular vein, which receives drainage from the inferior petrosal sinus, and the venous plexus of the hypoglossal canal outside the jugular foramen. The main structures in the area are the styloglossus, stylopharyngeal, and stylohyoid, the digastric nerve, and the stylomandibular ligament. The medial part of the temporal bone is constituted mainly by the internal auditory canal, the carotid canal, and the petrous apex (7,11). Laterally, within the petrous part of the temporal bone on the medial side of the mastoid antrum,
lies the semicircular canals and vestibule enclosed within the otic capsule (Fig. 5). The tympanic segment of the facial nerve passes below the lateral semicircular canal, and the mastoid segment descends to the stylomastoid foramen. The vestibule (vestibular cavity), which communicates with both ends of the semicircular canals, is situated medial to the lateral semicircular canal and below the superior semicircular canal. The aditus of the mastoid antrum opens into the tympanic cavity, which contains the malleus, incus, and stapes; the chorda tympani and tympanic nerve; the tensor tympani; and stapedius muscles. The tympanic cavity is limited laterally by the tympanic membrane, medially by the bone over the cochlea, and opens anteriorly into the eustachian tube. The arteries feeding the area arise from the stylomastoid, anterior tympanic, petrosal, and caroticotympanic arteries. Posterolateral to the otic capsule, anterior to the sigmoid sinus, and inferior to the
Chapter 1: Anatomy of the Cranial Base
A
33
B
Infraorb. N.
Infraorb. N.
Zygoma Infratemp. fossa Max. sinus Infratemp. fossa
Maxilla
C
D Nasolac. duct
Infraorb. canal Pteryg. Ven. Plex. Max. A. Max. sinus Post. wall
V2
Nasal septum
nasal Septum Pterygopal. fossa
Mid. concha Inf. concha Vomer
Lat. Pteryg. M.
Infratemp. fossa
Maxilla
Clivus
V3 Brs.
E
F
Sella Cav. Seg.
V2 Sphen. Septum Pterygopal. Fossa Pteryg. Ven. Plex.
Pterygopal. fossa
Cav. sinus
Max. A. V2 Pet. Seg. CN VI Bas. A.
Vert. A. V3 Brs. For. magnum
Figure 22 (A–C) Transmaxillary exposure of the cranial base. (A) In this dissection, a midfacial soft tissue flap has been reflected laterally to expose the anterior surface of the right maxilla. The operative approach to the maxillary sinus is more commonly performed using a sublabial incision in the gingivobuccal margin rather than through an incision on the face. The approach can be completed without dividing the infraorbital nerve, but in this dissection, it was divided below the infraorbital foramen. The nerve, if divided, can be resutured at the time of closing. The infratemporal fossa, which is situated below the greater sphenoid wing, has been exposed by removing the coronoid process of the mandible and a narrow wedge of zygoma. (B) The anterior wall of the maxillary sinus has been removed. The roof of the maxillary sinus forms the majority of the floor of the orbit. The infratemporal fossa contains the pterygoid muscles, mandibular nerve, maxillary artery, and the pterygoid venous plexus. (C) The medial and lateral walls of the maxillary sinus have been opened, but the posterior part of the sinus wall, which forms the anterior wall of the pterygopalatine fossa, has been preserved. Removing the medial wall of the sinus exposes the nasal cavity, turbinates, and nasal septum. The maxillary artery crosses the lateral pterygoid muscle to reach the pterygopalatine fossa, which is located behind the upper part of the posterior wall of the maxillary sinus and below the orbital apex. (D) The posterior wall of the maxillary sinus has been removed to expose the pterygopalatine fossa and orbital floor. The pterygopalatine fossa is located below the orbital apex and the posteromedial part of the inferior orbital fissure. The maxillary nerve enters the pterygopalatine fossa by passing through the foramen rotundum. The maxillary nerve gives rise to the infraorbital nerve, which passes forward in the infraorbital canal in the sinus roof and orbital floor. (E) Enlarged view of infratemporal and pterygopalatine fossae. Distally, the maxillary artery enters the pterygopalatine fossa, which is located in the lateral wall of the nasal cavity below the orbital apex. (F) The exposure has been directed medially through the nasal cavity to the clivus, which has been opened to expose the vertebral and basilar arteries and the front of the brain stem. The exposure has been extended upward by opening the sphenoid sinus and exposing the left intracavernous carotid. The margin of the foramen magnum has been preserved (1). Abbreviations: A., artery; Bas., basilar; Brs., branches; Cav., cavernous; CN, cranial nerve; For., foramen; Inf., inferior; Infraorb., infraorbital; Infratemp., infratemporal; Lat., lateral; M., muscle; Max., maxillary; Mid., middle; N., nerve; Nasolac., nasolacrimal; Pet., petrosal; Plex., plexus; Post., posterior; Pteryg., pterygoid; Pterygopal., pterygopalatine; Seg., segment; Sphen., sphenoid; Ven., venous; Vert., vertebral.
34
Martins and Rhoton
A
B Temp. M. CN VII Front. Br. Sup. Temp. A.
Lat. canthal Lig. Inf. Obl. M.
Zygoma
Infraorb. N. CN VII
Nasal cavity Max. sinus
Coronoid Proc.
Mass. M.
C
Pterion
D
Cav. sinus V1 Pterygopal. Fossa
Lat. Pteryg. M. Inf. Obl. M. Infraorb. N.
Gr. Palat. N. & A.
V2 Mid. Men. A. Eust. tube V3
Max. A.
Nasal septum
Nasal septum
E
F Front. N.
Lac. gland Lac. gland
Lat. Rec. M. CN IV Inf. Obl. M. Inf. Rec. M.
Sup. Obl. M. Lac. N.
CN III Cav. Sinus V1 V2 Vidian N. Inf. Obl. M.
Infraorb. N.
Lat. Rec. M.
Inf. Rec. M. CN III to Inf. Obl. M.
V3
Figure 23 Upper subtotal maxillotomy. Exposure obtained with mobilization of the upper part of the maxilla. (A) This approach uses paranasal, lower conjunctival, transverse temporal, and preauricular incisions. In the usual approach, the cheek flap is elevated as a single layer using subperiosteal dissection. In this dissection, the layers of the cheek flap were dissected separately to illustrate the structures in the flap. The facial muscles and branches of the facial nerve are exposed. The parotid gland has been removed. The frontal branch of the facial nerve crosses the mid portion of the zygomatic arch. If facial nerve branches are transected in the approach, they are tagged in preparation for re-approximation at closure. (B) A hemicoronal scalp incision and reflection of the temporalis muscle expose the lateral orbital rim. The cheek flap containing the facial muscles, branches of the facial nerve, parotid gland, and masseter muscle has been reflected inferiorly to the level of the maxillary attachment of the buccinator muscle. The orbital, maxillary, and zygomatic osteotomies have been completed and the lower half of the orbital rim; the anterior, medial, and lateral walls of the maxillary sinus; and the zygomatic arch have been reflected. The lower horizontal cut, located at Le Fort I level, extends above the apical dental roots and hard palate and along the inferior nasal meatus medially. The maxillotomy, at this stage, does not include the posterior maxillary wall or cross the greater and lesser palatine canals. The lateral nasal wall was included with the maxillotomy to expose the nasal cavity. The infraorbital nerve, which crosses the orbital floor, may be preserved for reconstruction. (C) The posterior wall of the maxillary sinus has been removed to expose the pterygopalatine fossa and the palatine nerves and arteries. The base of the coronoid process was divided, and the temporalis reflected downward to expose the lateral pterygoid muscle and maxillary artery in the infratemporal fossa. (D) A frontotemporal bone flap has been elevated, and the dura covering the frontal and temporal lobes and lateral wall of the cavernous sinus has been opened, and the temporal lobe has been elevated. The pterygoid muscles, the pterygoid process and plates, and the part of the middle fossa floor formed by the greater sphenoid wing have been removed to expose the nerves passing through the foramina rotundum and ovale. The eustachian tube is exposed behind the mandibular nerve and the middle meningeal artery. (E) Magnified view of the cavernous sinus, superior orbital fissure, and orbit. The oculomotor, trochlear, and ophthalmic nerves course through the lateral wall of the cavernous sinus. The ophthalmic nerve sends its branches along the upper part of the orbit. The maxillary nerve exits the foramen rotundum and passes through the pterygopalatine fossa, where it gives rise to the infraorbital nerve that courses along the floor of the orbit. The mandibular nerve passes through the foramen ovale and sends its branches through the infratemporal fossa. The vidian nerve passes forward in the vidian canal below the maxillary nerve to join the pterygopalatine ganglion in the pterygopalatine fossa. (F) Enlarged view of the orbital exposure. The lacrimal gland sits on the superolateral margin of the globe. The lacrimal nerve courses above the lateral rectus muscle. The inferior oblique muscle passes below the attachment of the inferior rectus muscle and upward between the globe and lateral rectus muscle to insert on the globe near the tendon of insertion of the superior oblique muscle (1). Abbreviations: A., artery; Br., branch; Cav., cavernous; CN, cranial nerve; Eust., eustachian; Front., frontal; Gr., greater; Inf., inferior; Infraorb., infraorbital; Lac., lacrimal; Lat., lateral; Lig., ligament; M., muscle; Mass., masseter; Max., maxillary; Men., meningeal; Mid., middle; N., nerve; Obl., oblique; Palat., palatine; Proc., process; Pteryg., pterygoid; Pterygopal., pterygopalatine; Rec., rectus; Sup., superior; Temp., temporal, temporalis.
Chapter 1: Anatomy of the Cranial Base
35
B
A
Mid. Men. A. V3 V2
Arc. Emin.
Gr. Pet. N. CN V Post. root
C
D Sup. canal
CN VII Laby. Seg. Vert. crest
.
Tens. Tymp. M.
Genic. Gang.
Pet. Car. A. Sup. Vest. N.
Modiolus
Inf. Vest. N.
Coch. N.
Clivus CN V
Inf. Pet. sinus Int. Ac. meatus
CN VII Meat. Seg.
E CN V CN III Bas. A. Post. Comm. A. CN IV Car. A. SCA Ant. Chor. A.
PCA
Figure 24 Middle fossa approach to the internal acoustic meatus. (A) The vertical line shows the site of the scalp incision and the stippled area outlines the bone flap bordering the middle fossa floor. (B) The dura has been elevated to expose the middle meningeal artery, the greater petrosal nerve, and the arcuate eminence. (C) The roof of the meatus has been opened to expose the superior and inferior vestibular, facial, and cochlear nerves. The vestibule and semicircular canals are located posterolateral and the cochlea is located anteromedial to the meatal fundus. In the middle fossa approach, for an acoustic neuroma, the cochlea and semicircular canal are not opened, as seen in this dissection illustrating the important structures which are to be avoided in opening the meatus. The vertical crest (Bill’s Bar) separates the facial and superior vestibular nerves at the meatal fundus. The superior and inferior vestibular nerves are located posteriorly and the facial and cochlear nerves anteriorly in the meatus with the cochlear nerve passing below the facial nerve to enter the modiolus. The labyrinthine segment of the facial nerve courses superolateral to the cochlea. (D) The bone of the petrous apex between the trigeminal nerve and the internal acoustic meatus has been removed to complete an anterior petrosectomy and to expose the inferior petrosal sinus and the lateral edge of the clivus. (E) The dura, exposed in the anterior petrosectomy and facing the posterior fossa and the tentorium, has been opened to expose the upper brain stem, oculomotor, trochlear, and trigeminal nerves and the basilar artery (7). Abbreviations: A., artery; Ac., acoustic; Ant., anterior; Arc., arcuate; Bas., basilar; Car., carotid; Chor., choroidal; CN, cranial nerve; Coch., cochlear; Comm., communicating; Emin., eminence; Gang., ganglion; Genic., geniculate; Gr., greater; Inf., inferior; Int., internal; Laby., labyrinthine; M., muscle; Meat., meatal; Mid., middle; Men., meningeal; N., nerve; PCA, posterior cerebral artery; Pet., petrous, petrosal; Post., posterior; SCA, superior cerebellar artery; Seg., segment; Sup., superior; Tens., tensor; Tymp., tympani; Vert., vertebral; Vest., vestibular.
36
Martins and Rhoton
A
B Sternocleidomast. M.
Sp. Henle Incus Mast. antrum deep Mastoid Tip Sup. Semicirc. canal Digastric M. Lat. Semicirc. canal Post. Semicirc. canal
Suprameat. Triang.
Suprameat. crest
CN VII Tymp. Seg. CN VII Mast. Seg. Digast. Groove Jug. Bulb Trautman’s Triang.
Mid. fossa dura Asc. Pharyg. A. Br. Longiss. Cap. M. Sinodural angle
D
C CN V motor root CN V AICA Nerv. Intermed. CN VII
Incus Chor. Tymp. N. CN IX CN VII Tymp. Seg. CN VII CN VII Mast. Seg. Stapes Laby. Seg. CN X CN VII Meat. Seg. Jug. bulb Sup. Vest. N. Inf. Vest. N. CN IX CN X CN V CN VIII Flocc. AICA Sig. sinus
CN VIII
Flocc.
F
E Malleus Cochlea Modiolus Scala Tympani Scala Vestibuli Chor. Tymp. N. CN V AICA CN VIII Coch. N. CN VII
Sig. sinus
Clivus Spiral Crest
Inf. Pet. Sinus AICA
CN VII Bas. A.
CN IX
CN VI
Jug. Bulb CN V
Flocc. Pons
CN V Motor Root Figure 25 Mastoidectomy, retrolabyrinthine, partial labyrinthine, translabyrinthine, and transcochlear approaches. (A) Right mastoid. The retroauricular flap and the sternocleidomastoid muscle have been reflected forward and the trapezius and underlying splenius capitus have been reflected backward to expose the mastoid and attachment of the longissimus capitus muscle. The posterior belly of the digastric muscle originates medial to the mastoid tip along the digastric groove. The spine of Henle is positioned at the posterosuperior margin of the external meatus, superficial to the deep site of the lateral semicircular canal and junction of the tympanic and mastoid segments of the facial nerve. The supramastoid crest, a continuation of the superior temporal line, is positioned at approximately the level of the upper margin of the transverse and sigmoid sinuses. The area below the anterior part of the supramastoid crest and behind the spine of Henle, called the suprameatal triangle, is positioned superficial to the mastoid antrum. The semicircular canals are positioned deep to the mastoid antrum. (B) The drilling has been extended to expose the middle fossa dura above, the sigmoid sinus posteriorly and the jugular bulb below. The superior, lateral, and posterior semicircular canals are located deep to the mastoid antrum and suprameatal triangle. The superior canal projects upward below the arcuate eminence. The posterior canal faces the posterior fossa dura. The lateral canal is positioned above the tympanic segment of the facial nerve. The facial nerve passes below the lateral canal and turns downward to form the mastoid segment. The dura between the sigmoid sinus and the semicircular canals, named Trautman’s triangle, faces the anterior surface of the cerebellum and cerebellopontine angle. A meningeal branch of the ascending pharyngeal artery passes through the jugular foramen and ascends in the dura of Trautman’s triangle. The jugular bulb is positioned medial to the cortical bone overlying the digastric groove. The sinodural angle is positioned at the junction of the sigmoid, transverse, and superior petrosal sinuses, and where the sigmoid sinus intersects the middle fossa dura. The short process of the incus points toward the tympanic segment of the facial nerve passing between the lateral semicircular canal and the stapes sitting in the oval window. The endolymphatic sac sits beneath the dura on the posterior surface of the temporal bone above and medial to the lower part of the sigmoid sinus. (Continued).
Chapter 1: Anatomy of the Cranial Base
superior petrosal sinus lies the presigmoid dura, referred to as Trautman’s triangle, under which the endolymphatic sac sits.
POSTERIOR CRANIAL BASE Endocranial Surface The posterior cranial base corresponds to the floor of the posterior fossa, an area around the foramen magnum. It is formed by the sphenoid, temporal, and occipital bones (Figs. 12, 13, and 14) (7,12–14). Medially, it is formed by the dorsum sellae, basilar (clival) portion of the occipital bone, and the foramen magnum. Laterally, the endocranial surface is formed by the posterior surface of the temporal and the occipital bones, with the petro-occipital fissure and the jugular foramen lying between the occipital and temporal bones. The endolymphatic sac, which sits beneath the dura, inferolateral to the internal acoustic meatus, is connected through the endolymphatic duct with the vestibule. The facial–vestibulocochlear nerve complex courses through the internal auditory canal (Fig. 15) (15). The arrangement of the nerves inside the meatus is as follows: the facial nerve, anterior and superior; the superior vestibular nerve, superior and posterior; the inferior vestibular nerve, inferior and posterior; and the cochlear nerve, anterior and inferior (Figs. 5 and 15) (7,11). The intermediate nerve courses with the eighth nerve adjacent to the brain stem and jumps to the seventh nerve at some point along the cisternal or the meatal segments of the facial nerve. The trigeminal nerve exits the posterior fossa by passing through the porus trigeminus, a dural ostium that is located between the superior petrosal sinus and the petrous apex along the posterior margin of the trigeminal impression. The subarcuate fossa, a depression lateral to the internal auditory canal, is pierced by the subarcuate artery, which ends in bone in the region of the superior semicircular canal (Fig. 15). The jugular foramen lies below the internal auditory canal between the petrous part of the temporal bone and the condylar part of the occipital bone (Figs. 12 and 14) (16). It is divided into a medially situated petrosal part, through which the inferior petrosal sinus passes, a laterally situated sigmoid part through which the sigmoid sinus passes, and an intermediately positioned intrajugular part through which the nerves pass. The hypoglossal canal is located below and medial to the jugular tubercle and above the middle third of the occipital condyles, which project downward along the anterior
37
half of the foramen magnum. The posterior condylar canal, located behind the condyle, conveys the posterior condylar vein, which connects the vertebral venous plexus with the sigmoid sinus. The inferior petrosal sinus courses along the petroclival fissure, connecting the posterior cavernous sinus and jugular bulb. The abducens nerve ascends and pierces the dura to course extradural through Dorello’s canal located between the petrosphenoid ligament and the upper edge of the fissure between the dorsum sella and the petrous apex, to enter the cavernous sinus. Below the foramen magnum, at the level of the atlanto-occipital joint, and anterior to the tectorial membrane, the cruciform, apical, and alar ligaments maintain the stability of the odontoid and craniocervical junction. The petrosal cerebellar surface, which faces the posterior surface of the temporal bone, the anterior surface of the brain stem, and the cerebellar peduncles, faces the endocranial surface of the posterior skull base (Fig. 16). The medulla, pons, and mesencephalon face the clivus. The surface of the medulla is divided longitudinally by the preolivary and postolivary sulci, with the pyramid in front and the inferior cerebellar peduncle behind the olive. The hypoglossal nerve arises along the preolivary sulcus and the glossopharyngeal, vagus, and accessory nerves arise near the retroolivary sulcus. The vestibulocochlear and facial nerves arise a few millimeters above the retro-olivary sulcus in the lateral part of the pontomedullary sulcus. The abducens nerve arises in the medial part of the pontomedullary sulcus and ascends behind the clivus. The trigeminal nerve arises from the anterolateral surface of the midpons with the rootlets which join to form the motor root arising around the superior third of the sensory root. The trochlear nerve arises below the inferior colliculus and passes forward along the pontomesencephalic sulcus in the quadrigeminal and ambient cisterns to enter the tentorial edge just behind the cavernous sinus. The oculomotor nerve crosses the interpeduncular cistern and pierces the roof of the cavernous sinus. Both vertebral arteries enter the cranium at the posterior edge of the occipital condyles and ascend to join in the midline, thus forming the basilar artery (Figs. 12 and 16). The main branches of the intradural vertebral artery are the posterolateral spinal artery, which supplies the posterior third of the medulla and upper spinal cord; the anterior spinal artery, which joins its mate from the opposite side on the anterior surface of the cord to supply the anterior two-thirds of the medulla and upper spinal cord; and the
←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− Figure 25 (Continued) (C) The retrolabyrinthine exposure has been completed and the dura has been opened to expose cranial nerves V to X in the cerebellopontine angle. The vestibulocochlear nerve has been depressed to expose the facial nerve and the nervus intermedius. The motor root of the trigeminal nerve is exposed superomedial to the main sensory root. The glossopharyngeal and vagus nerves are at the lower margin of the exposure. The flocculus protrudes from the foramen of Luschka behind the vestibulocochlear nerve. The anteroinferior cerebellar artery loops laterally between the facial and vestibulocochlear nerves. (D) The translabyrinthine approach has been completed to expose the vestibulocochlear and facial nerves in the internal acoustic meatus. The meatal and labyrinthine segments of the facial nerve are exposed proximal and the tympanic and mastoid segments are exposed distal to the geniculate ganglion. The dura of Trautman’s triangle has been opened to expose the trigeminal, glossopharyngeal, and vagus nerves in the cerebellopontine angle. The anteroinferior cerebellar artery loops laterally into the meatus before turning back toward the brain stem. The facial and superior and inferior vestibular nerves are exposed at the fundus of the meatus. The cochlear nerve is hidden anterior to the inferior vestibular nerve. (E) The greater petrosal nerve has been sectioned just distal to the apex of the geniculate ganglion and the facial nerve has been displaced posteriorly for removal of the cochlea in the transcochlear approach. The semicircular canals and vestibule, the end organs of the superior and vestibular nerves, have been removed. The incus has been removed but the malleus remains attached to the tympanic membrane. Drilling has been extended forward into the cochlea. The cochlear nerve enters the modiolus in the center of the spiral turns of the cochlea. (F) Removal of the cochlea opens the channel for removing the remainder of the petrous apex. The exposure extends to the lateral edge of the clivus and the inferior petrosal sinus. The basilar artery and anterior surface of the pons are at the deep end of the exposure. The abducens nerve passes behind the anteroinferior cerebellar artery and lateral to the basilar artery. Abbreviations: A., artery; AICA, anteroinferior cerebellar artery; Asc., ascending; Bas., basilar; Br., branch; Cap., capitis; Chor., chorda; CN, cranial nerve; Coch., cochlear; Digast., digastric; Flocc., flocculus; Inf., inferior; Intermed., intermedius; Jug., jugular; Laby., labyrinthine; Lat., lateral; Longiss., longissimus; M., muscle; Mast., mastoid; Meat., meatal; Mid., middle; N., nerve; Nerv., nervus; Pet., petrosal; Pharyng., pharyngeal; Post., posterior; Seg., segment; Semicirc., semicircular; Sig., sigmoid; Sp., spine; Sternocleidomast., sternocleidomastoid; Sup., superior; Suprameat., suprameatal; Triang., triangle; Tymp., tympani, tympanic; Vest., vestibular.
38
Martins and Rhoton
A
B
CN III
Sup. Pet. sinus V. of Labbe’
CN IV
Trautman’s triangle Otic capsule
S.C.A. Sup. Pet. V.
Sp. Henle CN V
Jug. bulb
Bas. A. CN VI CN VII CN VIII
AICA
PICA Vert. A. CN X
C
CN IX
D SCA
CN IV
SCA Marg. Br. CN V V. of Labbe’
Int. Ac. meatus Sup. canal
CN VII CN VIII
CN V Lat. canal CN VII CN VIII
Post. canal
Chor. Tymp. N. CN IX CN X
CN VII
CN VII CN IX CN X
Figure 26 Presigmoid approach. (A) The insert shows the temporo-occipital craniotomy and the mastoid exposure. The mastoidectomy has been completed and the dense cortical bone around the labyrinth has been exposed. The tympanic segment of the facial nerve and the lateral canal are situated deep to the spine of Henle. Trautman’s triangle, the patch of dura in front of the sigmoid sinus, faces the cerebellopontine angle. (B) Retrolabyrinthine exposure. The presigmoid dura has been opened and the superior petrosal sinus and tentorium divided, taking care to preserve the vein of Labb´e, which joins the transverse sinus, and the trochlear nerve, which enters the anterior edge of the tentorium. The abducens and facial nerves are exposed medial to the vestibulocochlear nerve. The PICA courses in the lower margin of the exposure with the glossopharyngeal and vagus nerves. The SCA passes below the oculomotor and trochlear nerves and above the trigeminal nerve. (C) The semicircular canals have been opened. The superior canal is located under the middle fossa’s arcuate eminence and the posterior canal is located immediately lateral to the posterior wall of the internal acoustic meatus. (D) The labyrinthectomy has been completed to expose the internal acoustic meatus (7). Abbreviations: A., artery; Ac., acoustic; AICA, anteroinferior cerebellar artery; Bas., basilar; Br., branch; Chor., chorda; CN, cranial nerve; Int., internal; Lat., lateral; Marg., margin; N., nerve; Pet., petrosal; PICA, posteroinferior cerebellar artery; Post., posterior; SCA, superior cerebellar artery; Sp., spine; Sup., superior; Tymp., tympani; V., vein; Vert., vertebral.
Chapter 1: Anatomy of the Cranial Base
A
39
B Occip. A.
Int. Car. A.
Post. Aur. N. Sup. nuchal line
Parotid Gl. Gr. Aur. N. Sternocleidomast. M.
Stylomast. A. CN VII
C1 Trans. Proc. Int. Jug. V.
Digastric M. Jug. bulb
Inf. Obl. M. Sup. Obl. M.
Semicirc. canals Sig. sinus
C
D CN VII
CN VII Incus
Chor. Tymp. N. CN VII Jug. Bulb
Lat. canal Sup. canal
Sig. Sinus
Pet. Car. A.
Mid. fossa Int. Jug. V.
Stapes Round window Lat. canal
Tympanic N.
Post. canal Jug. Bulb
F
E Symp. Tr. Sup. Laryn. N. CN XII CN X
Post. canal
Int. Car. A. CN VII
CN XII CN IX Int. Car. A. CN X, XI CN IX Intrajug. ridge C1 Trans. Proc. Intrajug. Proc. Inf. Pet. sinus
CN XI
Jug. bulb Med. wall
CN IX CN X, XI
Occip. A. Figure 27 (A–D) Postauricular exposure of the jugular foramen. (A) The C-shaped retroauricular incision (insert) provides access for the mastoidectomy, neck dissection, and parotid gland displacement. The scalp flap has been reflected forward to expose the sternocleidomastoid muscle and the posterior part of the parotid gland. (B) The more superficial muscles and the posterior belly of the digastric have been reflected to expose the internal jugular vein and the attachment of the superior and inferior oblique muscles to the transverse process of C1. A mastoidectomy has been completed to expose the facial nerve, sigmoid sinus, and the semicircular canals. (C) Enlarged view of the mastoidectomy. The jugular bulb is exposed below the semicircular canals. The chorda tympani arises from the mastoid segment of the facial nerve and passes upward and forward. The tympanic segment of the facial nerve courses below the lateral canal. (D) The external auditory canal has been transected and the middle ear structures have been removed, except the stapes, which remains in the oval window. The lateral edge of the jugular foramen has been exposed by completing the mastoidectomy, transposing the facial nerve anteriorly, and fracturing the styloid process across its base and reflecting it caudally. The petrous carotid is surrounded in the carotid canal by a venous plexus. (E) A segment of the sigmoid sinus, jugular bulb, and internal jugular vein has been removed. The lateral wall of the jugular bulb has been removed while preserving the medial wall and exposing the opening of the inferior petrosal sinus into the jugular bulb. Removing the medial venous wall exposes the portion of the glossopharyngeal, vagus, accessory, and hypoglossal nerves that are hidden deep to the vein. The main inflow from the petrosal confluens is directed between the glossopharyngeal and vagus nerves. (F) The medial venous wall of the jugular bulb has been removed. The intrajugular ridge extends forward from the intrajugular process and divides the jugular foramen between the sigmoid and petrosal parts. The glossopharyngeal, vagus, and accessory nerves enter the dura on the medial side of the intrajugular process, but only the glossopharyngeal nerve courses through the foramen entirely on the medial side of the intrajugular ridge (14). Abbreviations: A., artery; Aur., auricular; Car., carotid; Chor. Tymp., chorda tympani; CN, cranial nerve; Gl., gland; Gr., greater; Inf., inferior; Int., internal; Intrajug., intrajugular; Jug., jugular; Laryn., laryngeal; Lat., lateral; M., muscle; Med., medial; Mid., middle; N., nerve; Obl., oblique; Occip., occipital; Pet., petrosal, petrous; Post., posterior; Proc., process; Semicirc., semicircular; Sig., sigmoid; Sternocleidomast., sternocleidomastoid; Stylomast., stylomastoid; Sup., superior; Symp., sympathetic; Tr., trunk; Trans., transverse; V., vein.
40
Martins and Rhoton
posteroinferior cerebellar artery (PICA), which courses around the medulla and supplies the suboccipital cerebellar surface. The main branches of the basilar artery are the anteroinferior cerebellar artery (AICA), which passes around the pons and supplies the petrosal cerebellar surface and the nerves entering the internal auditory canal; the superior cerebellar artery (SCA), which encircles the midbrain and upper pons and supplies the tentorial cerebellar surface and dentate nucleus; and the posterior cerebral artery (PCA), which passes to the supratentorial area. The main venous drainage is by way of the petrosal veins, which join the superior and inferior petrosal veins emptying into the superior and inferior petrosal sinuses. The vascular and neural structures of the posterior fossa can be divided into upper, middle, and lower groups related to the three cerebellar arteries: the upper group is related to the SCA, which encircles the midbrain, courses on the superior cerebellar peduncle and within the cerebellomesencephalic fissure, passes below the oculomotor and trochlear nerves and above the trigeminal nerve, and supplies the tentorial cerebellar surface; the middle group is related to the AICA, which encircles the pons, courses on the middle cerebellar peduncle, dips in the cerebellopontine fissure passing by and sending branches to the facial and vestibulocochlear nerves, and supplies the petrosal cerebellar surface; and the lower group is related to the PICA, which encircles the medulla, passes near or between the rootlets of the lower four cranial nerves to course on the inferior cerebellar peduncle, dips into the cerebellomedullary fissure, and supplies the suboccipital cerebellar surface (Figs. 12 and 16) (17).
Exocranial Surface This portion is divided in central and lateral portions (Fig. 14) (14,18,19). The center portion is formed by the basal (clival) part of the occipital bone, which slopes upward from the foramen magnum and has the pharyngeal tubercle for the attachment of the superior pharyngeal constrictor on its lower surface, and the occipital condyles lateral to its lower portion at the anterolateral margin of the foramen magnum. The hypoglossal foramen, conveying the hypoglossal nerve, crosses above the middle one-third of the long axis of the condyle. The posterior condylar canal carries the posterior condylar vein interconnecting the vertebral venous plexus with the sigmoid sinus. Lateral to the condyle lies the jugular process of the occipital bone, which forms the posterior edge of the jugular foramen, connects the squamosal and basal parts of the occipital bone, and receives the attachment of the rectus capitus lateralis muscle posterior to the jugular foramen and jugular bulb. Two grooves lateral to the jugular process, a medial one for the occipital artery and a lateral one, the digastric groove, for the origin of the posterior belly of the digastric muscle, separate the jugular process from the mastoid process. The remaining muscles are the longus capitis attached to the lower clivus and the rectus capitis anterior attached in front of the occipital condyle. The nerves of the area are the hypoglossal and C1 nerves. The vertebral artery ascends through the C1 transverse process and crosses medially behind the superior articular pillar of the atlas or the atlanto-occipital joint to enter the dura (Figs. 17 and 18). The first segment of the vertebral artery ascends from the subclavian artery running up to the transverse foramen of C6. The second segment runs from C6 to C2, where the artery changes to a more lateral direction. The third segment ascends laterally to reach the C1 transverse foramen and turns medially and horizontally behind the atlanto-occipital joint
to course in the depths of the suboccipital triangle delimited by three muscles: the superior oblique extending from the occipital bone to the C1 transverse process, the inferior oblique extending from the C1 transverse process to the C2 spine, and the rectus capitus posterior major extending from the C2 transverse process to the occipital bone (Figs. 17 and 18). The fourth segment of the vertebral artery extends from the dural entrance to the vertebrobasilar junction. The third segment gives rise to the posterior meningeal and muscular arteries and occasionally the PICA. The other vascular structures in the area are the occipital and ascending pharyngeal arteries, the vertebral venous plexus, and the posterior and anterior condylar veins.
DISCUSSION With the development of microsurgical techniques and skull base surgical principles, it has become possible to access all parts of the cranial base. Lesions involving the central part of the anterior two-thirds of the cranial base and clivus can be accessed through intracranial routes, such as the orbitozygomatic, transcranial–transbasal, or extended frontal approaches, or by subcranial routes utilizing the various modifications of the transnasal, transoral, transsphenoidal, transmandibular, transmaxillary, transcervical, or facial translocation approaches or by a combination of the intracranial and subcranial routes (Figs. 19–23). The approaches can also be extended to the middle skull base by using the orbitozygomatic, preauricular infratemporal fossa, subtemporal anterior petrosectomy, or other extensions of the middle fossa routes (Fig. 24). Further posteriorly, approaches directed through the temporal bone, such as the retrolabyrinthine, translabyrinthine, transcochlear, and combined supra-infratentorial presigmoid approaches, or a combination of preauricular and postauricular transtemporal approaches, may be considered (Figs. 25 and 26). Lesions in the posterior fossa and posterior skull base may be reached through the retrosigmoid or suboccipital routes or the farlateral approach and its transcondylar, supracondylar, and paracondylar modifications (Fig. 24) (19). The jugular foramen is most often accessed by a postauricular transtemporal approach (Fig. 27). Cranial base tumors frequently invade intracranial and subcranial spaces and require innovative combinations of these transcranial, subcranial, and combined approaches. Thoughtful consideration of skull base anatomy is essential to successful surgery for these tumors. REFERENCES 1. Rhoton AL Jr. The anterior and middle cranial base. Neurosurgery. 2002;51(1 Suppl):S273–S302. 2. Rhoton AL Jr., Seoane E. Surgical anatomy of the skull base. In: Harsh, G., ed. Chordomas and Chondrosarcomas of the Skull Base and Spine. New York: Thieme Medical Publishers Inc, 2003:57–79. 3. Rhoton AL Jr. The orbit. Neurosurgery. 2002;51(1 Suppl):S303– S334. 4. Rhoton AL Jr., Natori Y. The Orbit and Sellar Region: Microsurgical Anatomy and Operative Approaches. New York, NY: Thieme Medical Publishers Inc, 1996:3–25. 5. Natori Y, Rhoton AL Jr. Transcranial approach to the orbit: Microsurgical anatomy. J Neurosurg. 1994;81:78–86. 6. Natori Y, Rhoton AL Jr. Microsurgical anatomy of the superior orbital fissure. Neurosurgery. 1995;36:762–775. 7. Rhoton AL Jr. The temporal bone and transtemporal approaches. Neurosurgery. 2000;47(3 Suppl):S211–S265.
Chapter 1: Anatomy of the Cranial Base 8. Rhoton AL Jr. The sellar region. Neurosurgery. 2002;51(1 Suppl): S335–S374. 9. Rhoton AL Jr. The cavernous sinus, the cavernous venous plexus, and the carotid collar. Neurosurgery. 2002;51(1 Suppl): S375– S410. 10. Seoane E, Rhoton AL Jr., de Oliveira E. Microsurgical anatomy of the dural collar (carotid collar) and rings around the clinoid segment of the internal carotid artery. Neurosurgery. 1998;42:869– 886. 11. Pait TG, Zeal AA, Harris FS, et al. Microsurgical anatomy and dissection of the temporal bone. Surg Neurol. 1977;8:363– 391. 12. Rhoton AL Jr. Cerebellum and fourth ventricle. Neurosurgery. 2000;47(3 Suppl): S7–S27. 13. Rhoton AL Jr. The foramen magnum. Neurosurgery. 2000;47(3 Suppl):S155–S193.
41
14. Rhoton AL Jr. Jugular foramen. Neurosurgery. 2000;47(3 Suppl): S267–S285. 15. Rhoton AL Jr. The cerebellopontine angle and posterior fossa cranial nerves by the retrosigmoid approach. Neurosurgery. 2000;47(3 Suppl):S93–S129. 16. Katsuta T, Rhoton AL Jr., Matsushima T. The jugular foramen: Microsurgical anatomy and operative approaches. Neurosurgery. 1997;41:149–202. 17. Rhoton AL Jr. The cerebellar arteries. Neurosurgery. 2000;47(3 Suppl):S29–S68. 18. Rhoton AL Jr. The far-lateral approach and its transcondylar, supracondylar, and paracondylar extensions. Neurosurgery. 2000;47(3 Suppl):S195–S209. 19. Wen HT, Rhoton AL Jr., Katsuta T, et al. Microsurgical anatomy of the transcondylar, supracondylar, and paracondylar extensions of the far-lateral approach. J Neurosurg. 1997;87:555–585.
2 Pathology of Tumor and Tumor-like Lesions of the Skull Base Michelle D. Williams and Adel K. El-Naggar
INTRODUCTION
to avoid complications such as CSF leak. Generally, these lesions present as smooth, homogenous, tan soft tissue mimicking brain parenchyma. Histologically, they are typically composed of neural tissue with fibrosis and astrocytic and gametocytic cell proliferation.
The skull base regions are composed of complex tissue structures that give rise to histogenetically and biologically heterogeneous neoplasms of ectodermal, endodermal, and mesodermal origins. The morphologic and histogenetic differences are complicated by overlapping features and frequently pose diagnostic difficulties. The vast majority of tumors at these locations are malignant with a small percentage being benign or tumor-like lesions (Table 1). Accurate diagnosis and understanding of the clinical and pathologic presentations of the varied tumor entities in this region are essential for proper management. Although the majority of tumors at this location are of primary origin, metastasis can be encountered and will be discussed. Nonmetastatic tumors are either primary or an extension from neighboring structures.
Differential Diagnosis This entity can be differentiated from encephalocele, which frequently shows meninges.
Respiratory Epithelial Adenomatoid Hamartoma This is a benign proliferation of minor seromucinous glands of the sinonasal tract, occurring more commonly in men than women in their sixth decade of life. The major symptoms include nasal obstruction, epistaxis, and recurrent sinusitis. These lesions appear normally as polypoid tan to reddishbrown and rubbery tissue nodules (4–6). Histologically, they are formed of numerous glandular structures lined by ciliated respiratory epithelium with thickened basement membrane and intervening fibrotic and/or edematous stroma (Fig. 2).
Biopsies and Frozen Sections The evaluation of sinonasal pathology typically requires a tissue biopsy that may be limited dependent upon the accessibility of the target region. Obtaining adequate and representative materials is essential for accurate diagnosis and better planning of patient management. The initial assessment of these tumors is commonly conducted intraoperatively for either provisional or definitive diagnosis and/or verifying adequacy for representative tissue. Communication with the pathologist at the time of frozen section is key to coordinating patient care. At this stage, non-neoplastic processes, lymphoma, and metastatic neoplasms can be determined. For primary tumors, the frozen tissue biopsy may be adequate for diagnosis and planning ancillary tests but efforts to secure additional tissue for permanent processing is strongly recommended for optimal morphologic assessment and biomarker characterization.
Differential Diagnosis These lesions may be confused with Schneiderian inverted papilloma and sinonasal adenocarcinomas. The benign glandular structures lined by columnar cells that form these lesions are key to differentiating it from both of these entities.
Ectopic Pituitary Adenoma Pituitary adenomas may occur in the sphenoid bone and sinuses either as a separate lesion or as an extension from a primary adenoma arising in the sella (7,8). Embryonic residue along the Rathke pouch formation is the presumed derivation. Females are more affected than males (2:1 ratio). Patients may present with nasal obstruction, headache, or epistaxis. Approximately half of patients manifest hormonal abnormalities. Histologically, an ectopic pituitary adenoma is identical to that of a conventional pituitary adenoma with monotonous round cells (Fig. 3).
NON-NEOPLASTIC AND CONGENITAL LESIONS Encephalocele Based on the age of the patient and the location of the lesion, the diagnosis of an encephalocele may be made by imaging prior to submitting histology. Frequently, encephaloceles extending into the nasal cavity or sinus include meninges and glial tissue associated with fibrosis (Fig. 1) (1,2).
Differential Diagnosis This lesion should be differentiated from carcinoid tumor, neuroblastoma, and other small undifferentiated tumors at these locations. Immunohistochemical staining for hormonal receptors, especially for ACTH and prolactin, is helpful.
Nasal Glial Heterotopia This is a congenital malformation in which ectopic glial tissue is found without connection to intracranial structures (3). Nasal glial heterotopia may present as an extranasal, intranasal, or mixed intranasal and subcutaneous mass. Patients may also present with symptoms and findings of nasal polyp, chronic sinusitis, and otitis media. Radiological confirmation of the lack of intracranial communication is stressed
Inflammatory Pseudotumor This is a benign reactive process where a spindle cell tumor– like proliferation with inflammatory component is the cardinal feature (9–11). They may arise at any site in the skull base regions. 43
44
Williams and El-Naggar Table 1
Incidence of Malignant Sinonasal Tumors∗
Feature
%
Sites • Maxillary • Nasal cavity • Ethmoid • Frontal & sphenoid
60% 22% 15% 3%
Derivation • Epithelial • Mesenchymal • Neuroectodermal • Others
55% 30% 15% 5%
∗ ∗
<1% of all neoplasms or tumors 3% of head and neck malignancy
Figure 3
Pituitary adenoma with uniform endocrine cells.
Histopathology These tumors are composed of spindle cell proliferation admixed with chronic inflammatory and plasma cells. Immunohistochemically, the spindle cells express weak smooth muscle actin (Fig. 4) characteristic of myofibroblast.
Differential Diagnosis These lesions may be confused with some benign and malignant mesenchymal tumors and fibromatosis. The inflammatory cells and the myofibroblastic nature are keys to their proper identification.
Sinonasal Polyps
Figure 1 Encephalocele with glial brain tissue (arrow).
These are polypoid growth originating from the Schneiderian epithelial lining of the sinonasal cavities. They evolve as a result of fluid accumulation of the mucosa, most commonly due to nasal allergy and repeated sinusitis. Frequently, maxillary polyps may extend via the sinus openings into nasal or nasopharyngeal locations. The majority of polyps
Figure 2 Adenomatoid hamartoma composed of respiratory epithaliel lined spaces and glandular structures.
Figure 4 Inflammatory pseudotumor with admixed spindled and inflammatory cells.
Chapter 2: Pathology of Tumor and Tumor-like Lesions of the Skull Base
45
Figure 5 Vasular sinonasal polyp with scattered inflammatory cells and infammatory edematous stroma. Figure 6 Cylindrical cell papilloma with characteristic bland columnar epithelial lined papillae and microcystic formation.
occurring in children are associated with asthma or cystic fibrosis (20–30%). In adult patients, they are often secondary to chronic sinusitis and allergy (12–14). Generally, sinonasal polyps grossly appear smooth, glistening, and translucent or opaque. Histologically, the epithelium of nasal polyps is commonly respiratory with and without mild hyperplasia or squamous metaplasia. The core is edematous with numerous vessels and with scattered inflammatory cells. In allergic polyps, the dominant cells are eosinophils, whereas in inflammatory polyps, there are chronic and a few acute inflammatory cells. Scattered minor salivary gland structures are also seen. Commonly, secondary changes due to infarction and organization are observed (Fig. 5).
Exophytic Schneiderian Papilloma Generally, exophytic Schneiderian papilloma is a lobular mass composed histologically of bland squamous epithelium around fibrovascular cores.
Cylindrical Cell Papilloma A generally exophytic papillomatous proliferation, histologically lined by multilayered columnar epithelium with and without oncocytic features. Microcysts with neutrophils within the epithelium are common (Fig. 6).
Inverted Schneiderian Papilloma
Differential Diagnosis These lesions should be differentiated from angiofibroma, and small cell tumors forming a polypoid mass. Angiofibroma is a specific benign entity with proliferation of vascular spaces with satellite bland stromal cells in young male patients. Rhabdomyosarcoma, lymphoma, and melanoma are more cellular and can be readily excluded using respective markers.
Schneiderian Papillomas There are three histologic subtypes of this entity: (i) the exophytic which is typically squamous in histology and almost always affecting the nasal septum, (ii) cylindrical cell is characterized by stratified columnar epithelial lining, and (iii) the inverted papilloma is where a transitional-like squamous proliferation in an inward growth constitutes its cardinal feature (15–20). The latter types originate from the lateral nasal wall, middle meatus, and paranasal sinuses. Both cylindrical and inverted papillomas affect one side of the nasal and paranasal sinuses with less than 5% occurring bilaterally.
Grossly, inverted papilloma are typically tan to gray, polypoid soft tissue with a mulberry appearance. These lesions affect more males than females and occur in older age. Keratinization may be present in some lesions and this feature may indicate potential progression. A characteristic feature of this lesion is inward invagination and cellular structures formed of transitional-like epitholium with a intraepithelial microcysts filled macrophages, cellular debris, and mucin like materials. Malignant transformation may occur in 10% of these lesions. Identifying dysplasia and carcinoma in situ in these lesions is critical to predict the malignant progression of these lesions (Fig. 7).
Differential Diagnosis All three types of Schneiderian papillomas must be differentiated from carcinomas: the exophytic type from papillary squamous carcinomas, cylindrical cell papilloma from sinonasal adenocarcinomas, and the inverted type primarily from squamous carcinoma, which will show dyskeratosis, dysplasia, and stromal desmoplasia in response to invasion.
Human Papillomavirus Low- and high-risk human papillomavirus (HPV) subtypes have been identified in inverted and exophytic papillomas by in-situ hybridizations and polymerase chain reactions. No clear association between HPV status and malignant transformation has been established.
INFLAMMATORY AND GRANULOMATOUS CONDITIONS Allergic Fungal Sinusitis This is a specific entity resulting from an allergic reaction to persistent fungal organisms leading to mass
46
Williams and El-Naggar
Figure 7 Inverted Schneiderian papilloma composed of an epithelial proliferation within submucosa.
Figure 9 Invasive fungal sinusitis with fungal hyphae invading vasculature with tissue necrosis. A silver stain highlights the fungal hyphae (inset).
Differential Diagnosis formation (21–23). This may occur at any age, without gender predominance. The main symptoms are nasal discharge and rhinorrhea over a protracted period. The overall appearance of tissues from these lesions is that of tan, butter-like material with debris. The cardinal microscopic features include characteristic alternating layers of mucin and degenerating cells with eosinophilic materials and cells (Fig. 8). Numerous intact and degenerating eosinophils and granules are present. Occasionally, crystalloid eosinophilic deposition (Charcot–Leyden) crystals are found. Tissue invasion by hyphae is not identified in this disorder.
Ancillary Markers Fungal stains may reveal degenerated non-invasive fungal in allergic mucin secretion. Culture is necessary for speciation of the fungal organism. The most common species is Aspergillus.
The most common differential diagnosis of this entity is invasive fungal infection, mycetoma (fungal ball) or inflammatory polyp. Invasive fungal infections are characterized by vascular invasion by fungal organism and tissue necrosis and the lack of layered secretions. Mycetoma shows sheets of fungal hyphae without tissue invasion. Inflammatory polyp can be excluded based on the lack of alternating layers and eosinophilic materials and cells.
Invasive Fungal Sinusitis Invasive fungal sinusitis is a potentially life-threatening condition characterized by fungal organisms invading tissue and blood vessels leading to massive tissue necrosis (24,25). Most commonly it is seen in immunocompromised patients and patients with poorly controlled diabetes mellitus. Infections spread rapidly to involve the central nervous system, which may lead to patient’s death. Normally, tissue fragments are tan-gray, soft, and necrotic. Histologically, fungal hyphae are identified invading tissues and specifically the blood vessels leading to massive necrosis (Fig. 9). Zygomycoses (Rhizopus/Rhizomucor) and Aspergillus are the most common organisms. Cultures should be taken for definitive species identification.
Differential Diagnosis The most common differential diagnoses of this entity are allergic fungal sinusitis and infectious agents. Invasive fungal infections are characterized by vascular invasion by fungal organism and tissue necrosis and the lack of layered secretions. Inflammatory polyp can be excluded based on the lack of alternating layers and eosinophilic materials and cells.
Rhinoscleroma
Figure 8 Allergic fungal sinusitis is hallmarked by Allegic mucin dominated by eosrnophils.
Rhinoscleroma is a chronic progressive granulomatous inflammation caused by Klebsiella rhinoscleromatis, a gramnegative bacterium (26,27). The disease is rare in the United States but is endemic in developing counties. Females are slightly affected more than males in their second to third decades of life. The natural history of the disease includes exudative, proliferative, and fibrotic phases. The exudative stage is characterized by purulent discharge, acute and
Chapter 2: Pathology of Tumor and Tumor-like Lesions of the Skull Base
Figure 10 Rhinoscleroma; inflammatory infiltrate including the characteristic vacuolated histiocytes (Mikulicz cells).
chronic inflammation, and mucosal edema and congestion. The proliferative phase is characterized by granulomatous formation and multiple nodular inflammatory and ulcerating masses. The terminal phase is fibrotic, leading to stenosis. The overall appearance of this lesion is that of polypoid and friable soft tissue in the proliferative phase. The histologic features are those of chronic inflammatory conditions with extensive histiocytic cells with vacuolation containing Klebsiella microorganisms, (Mikulicz cells) (Fig. 10). The fibrotic stage is nondiagnostic of the condition.
Differential Diagnosis Other infectious agents including mycobacterial infections and syphilis also need to be considered.
47
Figure 11 Histiocytosis-X (eosinophilic granuloma), characteristic foamy histiocytes and eosinophils are shown.
Histiocytosis-X (Eosinophilic Granuloma) Histiocytosis-X (eosinophilic granuloma) is the most prevalent of these rare conditions, which also include two related ¨ conditions Hand-Schuller-Christian syndrome and Lettere– Siwe disease. This disease can be localized or systemic and frequently involve head and neck sites including the flat bones of the skull and sinonasal tract. Although age at presentation varies widely, it is most common in younger groups (<20 years). Symptoms are nonspecific and include middle ear infections and destructive bony structures (32–36). The cardinal histologic features are proliferation of large histiocytes with cytoplasmic vacuolation and marked eosinophilic infiltrate (Fig. 11). The characteristic electron dense granules, Birbeck, are often seen on electronoptic examinations.
Differential Diagnosis Wegener Granulomatosis A destructive granulomatous condition of unknown etiology. The disease, although systemic in nature, most commonly affects the upper respiratory tract. The disease affects all ages and both genders. Common complaints are nasal stuffiness, rhinitis, and pain. Nasal septal destruction is common in young patients (28–31). Generally, tissues obtained from these lesions are composed of fragments of necrotic appearing materials. Histologically, the characteristic features include small vessel inflammation with granulomatous features, geographic necrosis, and basophilic debris due to cellular degeneration.
Ancillary Markers Up to 85% of patients with Wegener granulomatosis are positive for serologic testing with antineutrophil cytoplasmic antibodies (c-ANCA).
Differential Diagnosis All infections associated with granulomatous inflammation would be included in the differential diagnosis, and should be excluded by negative cultures. NK/T-cell lymphoma may show necrosis and may be included.
The conditions most frequently confused with this entity are Hodgkin disease and lymphoma NK/T-cell type.
Myospherulosis This is an iatrogenically induced lesion caused by reaction to petroleum, lanolin-based products, and fat necrosis (37). Typically, patients present having undergone surgery with petroleum-based nasal packing. Patients’ symptoms include sinusitis, pain, and swelling. The main histologic findings are pseudocystic formations containing nonrefractile small spherules and fibrosis.
Differential Diagnosis Coccidioidomycosis fungal infection.
MALIGNANT EPITHELIAL NEOPLASMS Sinonasal malignancy most commonly affects the maxillary sinus (60%), the nasal cavity (22%), ethmoid sinus (15%), and less frequently the frontal and sphenoid sinuses (3%) (Table 1). A wide variety of malignant neoplasms of different cellular lineages arises in this location. The most common
48
Williams and El-Naggar Table 2 Clinicopathologic Features of Undifferentiated Carcinoma of the Skull Base Feature
SNUC
NEC
NPC-type
Grade Incidence M/F LN Mets Mortality Risk factor Site
High Rare 3:1 30% 80% NI Nasal Cavity & sinus
High Rare NI NI 50–60% NI Maxillary sinus
High <0.5% 3:1 Common 50–60% EBV Nasopharynx
Abbreviations: EBV, Epstein–Barr virus: LN Mets, Lymph node metastases; M/F, Male/Female; NEC, neuroendocrine carcinoma; NPC, nasopharyngeal carcinoma; SNUC, sinonasal undifferentiated cancer; NI, no information.
Figure 12 Squamous carcinoma, keratin, forming tumor nest formation with invasive growth pattern.
Histologically, SNUC is composed of sheets of undifferentiated cells with low nuclear-cytoplasmic ratio, prominent nucleoli and reticular and clear nuclei with high mitotic rate and necrotic features (Fig. 13). No squamous or glandular differentiation should be present.
Differential Diagnosis malignancies are carcinomas (55%) followed by nonepithelial tumors (30%), neuroectodermal (15%), and miscellaneous entities (5%) (38,39).
Squamous Carcinoma Squamous carcinoma typically occurs in elderly individuals in their sixth and seventh decades with a tendency for male gender (M:F, 2:1). The most common sites are maxillary sinus, nasal cavity, ethmoid, frontal, and sphenoid sinus (40–43). Usually, these carcinomas are typical ulcerated and indurate with exophytic features, sinus tumors are bulky and composed of friable light tan tumor tissues. Histologically, squamous features should be identified and the extent of differentiation graded (well, moderate, or poor). Tumors originating from the nasal cavity are generally keratinizing and well differentiated (Fig. 12). Tumors of the paranasal sinuses and skull base are nonkeratinizing carcinomas with intermediate to poor differentiation. The poorly differentiated carcinomas are also called Schneiderian carcinoma and most likely originate from preexisting inverted papillomas.
The diagnosis and etiology of this entity is a subject of controversy and frequently arise as management issues. Because of the poor differentiation, they are most often confused with poorly differentiated sinonasal carcinomas (Schneiderian/squamous) as described above, neuroendocrine carcinoma, and the solid form of adenoid cystic carcinoma. Immunohistochemical markers are crucial to the diagnosis of this entity. Positive epithelial lineage markers and absence or focal neuroendocrine marker exclude neuroblastoma, neuroendocrine carcinoma, lymphoma, melanoma, and primitive neuroectodermal tumor (PNET). The lack of lymphoid infiltrate along with accurate localization is crucial for the exclusion of nasopharyngeal-type carcinoma (Table 2).
Nasopharyngeal Carcinoma They are classified into keratinizing and nonkeratinizing phenotypes and correspond to WHO type I and II/III grades (50–56). Pathogenesis is strongly linked to Epstein–Barr virus
Differential Diagnosis Squamous carcinoma may occasionally pose a diagnostic challenge on small biopsy specimens with sialometaplasia, pseudoepitheliomatous hyperplasia, and Schneiderian papilloma and should be differentiated from sinonasal undifferentiated carcinoma.
Sinonasal Undifferentiated Carcinoma Sinonasal undifferentiated carcinoma (SNUC) is a highgrade, undifferentiated carcinoma characterized by primitive malignant epithelial cells with high mitotic figures and cellular necrosis (44–49). Patients are typically elderly males who present in a late stage of disease (Table 2). Tumors originate most commonly from the nasal cavity, ethmoid sinus, and maxillary sinus but site of origin may be difficult to certain at presentation. Generally, these tumors are typically large, involving multiple adjacent structures with ill-defined borders. They present as light-tan, soft tissue fragments.
Figure 13 SNUC, undifferentiated neoplastic cells with prominent nucleoli and necrosis.
Chapter 2: Pathology of Tumor and Tumor-like Lesions of the Skull Base
Figure 14 Nasopharyngeal carcinoma, undifferentiated type (WHO-III) showing undifferentiated neoplastic cells with cleared nuclei in a prominent lymphoid background.
(EBV) infection and a diet high in nitrosamines including salted fish and fermented food (Table 2). The histopathologic characteristics of these tumors vary according to the WHO classification. Keratinizing type (WHO-I) is composed of tumor nests with squamous features including keratinization and intercellular bridges. Nonkeratinizing type (WHO-II/III) are composed of sheets of undifferentiated malignant epithelial cells intimately intermingled with chronic inflammatory infiltrate, which is often EBV positive (Fig. 14).
Ancillary Markers On histologic examination, EBV can be identified by in-situ hybridization for EBV-encoded RNA. Most commonly, EBV is identified in nonkeratinizing types (WHO-II/III). Serology for EBV-encoded RNAis also available.
Differential Diagnosis The nonkeratinizing phenotype must be differentiated mainly from the SNUC and neuroendocrine carcinoma. The presence of lymphocytic infiltrate and the lack of neuroendocrine markers helps to confirm the diagnosis.
Sinonasal Adenocarcinoma These tumors originate from either the respiratory epithelium or the underlying seromucinous glands. Tumors of respiratory epithelial derivation are frequently located in the nasal cavity and the ethmoid sinus, while those arising from the subepithelial glands frequently affect the nasal cavity and the maxillary sinus. Generally, salivary-type tumors appear as large, nodular, light tan soft masses, whereas the intestinal-type adenocarcinomas tend to be bulky, friable, and soft with ulcerated surface.
Salivary-Type Adenocarcinomas The most frequent phenotype affecting minor salivary gland of the skull-base is adenoid cystic carcinoma. Less frequent subtypes are mucoepidermoid, acinic cell, and low-grade papillary adenocarcinoma (57,58).
49
Figure 15 Adenoid cystic carcinoma, cribriform type.
The histopathologic features are those of their primary salivary counterparts (Fig. 15).
Differential Diagnosis The differential diagnosis is mainly of the solid form of adenoid cystic carcinoma, which should not be confused with basaloid squamous or neuroendocrine carcinomas. The lack of squamous differentiation, neuroendocrine markers, and anaplastic features favors adenoid cystic carcinoma.
Nonsalivary Gland Adenocarcinoma Tumors under this category are divided into intestinal and nonintestinal (seromucinous) adenocarcinomas. The intestinal adenocarcinoma is identical to those arising in the intestinal tract. These tumors arise in patients with history of exposure to hardwoods, leather, and certain chemical manufacturing. These tumors tend to occur in the ethmoid sinus and nasal cavity. The nonintestinal adenocarcinomas are typically seromucinous adenocarcinoma and affect the ethmoid and maxillary sinuses (59–64). Histologically, the intestinal type is typically that of colonic adenocarcinoma phenotype but may show mucinous and signet-ring features (Fig. 16). The seromucinous type is usually low grade with back-to-back cuboidal-lined glands and cords (Fig. 17).
Differential Diagnosis The intestinal form should be differentiated from metastasis to the skull base areas from salivary and intestinal primaries, which should be clinically excluded (Table 3).
MESENCHYMAL TUMORS Benign Tumors Various benign tumors similar to those originating in other soft tissues may arise in the skull base region.
Lobular Capillary Hemangioma (Pyogenic Granuloma) This is a relatively common benign vascular lesions representing approximately 25% of the nonepithelial neoplasms of the
50
Williams and El-Naggar
Figure 16 Sinonasal adenocarcinoma, nonsalivary, intestinal type, forming neoplastic glands with large pseudostratified nuclei.
Figure 18 Pyogenic granuloma with a lobular proliferation of capillaries in the stroma.
sinonasal tract and skull base region. This entity frequently found is the nasal septum (60%) and is most commonly identified in adolescent boys and young women. Local trauma and hormonal factors may play an etiologic role in the development of this entity. Intermittent painless epistaxis is the most common symptom (65,66). Usually, a polypoid red to purple nodular growth is identified with mucosal ulceration. Histologically, lobular capillary hemangiomas are composed of fairly organized vascular proliferations with lobular formation (Fig. 18). An inflammatory infiltrate may be present.
ture is the key finding of capillary hemangioma, which will lack the stellate stroma of an angiofibroma or the highly cellular spindled cells of a hemangiopericytoma.
Differential Diagnosis The differential diagnoses of these lesions are mainly from vascular polyp, nasopharyngeal angiofibroma, hemangiopericytoma, and low-grade angiosarcoma. The lobular architec-
Figure 17 Sinonasal adenocarcinoma, nonsalivary, nonintestinal (seromucinous) type, formed by bland cuboidal back-to-back glands filling the stroma.
Hemangiopericytoma An uncommon nonepithelial neoplasm is composed of spindle cell proliferation of hybrid pericyte and myxoid differentiation. The tumor can present at any age, and patients most commonly complain of nasal obstruction epistaxis and pain. The nasal cavity and the paranasal sinuses are typically affected (67–70). In general, these tumors appear typically as a polypoid gray to red, soft, fleshy tissue mass. Histologically, these lesions manifest subepithelial spindled to round markedly compacted cell proliferation with complex vascularity (Fig. 19). The cells may form a variety of patterns including fascicular, storiform, and palisading morphology with interspersed vascular spaces in different
Figure 19 Hemangiopericytoma showing proliferation of spindled cells with intervening staghorn vascular spaces.
Chapter 2: Pathology of Tumor and Tumor-like Lesions of the Skull Base Table 3
51
Sinonasal Adenocarcinomas Nonsalivary
Parameter
Salivary
Intestinal
Seromucinous type
Origin Age (years) Gender Prognosis Recurrence Risk factor
Minor salivary gland 30–70 Equal Depends on stage 50% survival High (60%) unknown
Minor salivary gland 30–70 Equal • Depends on stage • 50% Yes unknown
Markers
S-100, Keratins, SMA
Respiratory mucosa 60–70 More in males • Depends on differention • and stage High • Wood worker • Leather worker K-ras, Keratins 7 and 20
size and forms; mitosis and mild cellular pleomorphism may also be seen.
Ancillary Markers By immunohistochemistry, the spindled cells are positive for smooth muscle actin and negative for CD34, which will highlight the vessels.
Differential Diagnosis This tumor should be readily differentiated from reactive pyogenic granuloma, cellular hemangioma and nasopharyngeal angiofibroma, benign and low-grade smooth muscle tumors, solitary fibrous tumor, and spindle cell sarcomas (synovial sarcoma and fibrosarcoma). Immunohistochemical markers may be used to exclude some of these entities. In solitary fibrosis tumors, there is lower cellularity and positivity for CD34 and bcl-2 in the spindled cells, which is helpful in the differentiation of these entities (73–75).
Nasopharyngeal Angiofibroma A benign highly vascular mesenchymal neoplasm arising predominantly in young males. The lesions arise in the roof of the nasopharynx. Patients’ symptoms typically include nasal obstruction, epistaxis and drainage, less commonly, facial deformities, proptosis, deafness, sinusitis, and palliative swelling (73,75). Usually, angiofibroma presents as rounded, nonencapsulated, gray-white soft tissue masses covered with smooth mucosa and spongy cut surface appearance. The cardinal histologic features of these lesions are richly thin-walled vascular formations in fibrotic connective background with stromal cell proliferation (Fig. 20).
S-100, Keratins, SMA
Histologically, they are characterized by sparsely scattered, stellate-shaped mesenchymal cells in a myxoid background.
Malignant Neoplasms Rhabdomyosarcoma This is a relatively uncommon mesenchymal malignancy of the skull base region. Rhabdomyosarcoma is the most common sarcoma of the head and neck and is the most frequent childhood sarcoma. The sinonasal tract and the nasopharynx are the most commonly affected sites. The embryonal type is the most common type in children while the alveolar subtype predominates in an older age group. Patients present with swelling, bleeding, visual symptoms, and sinusitis (Table 4) (78–82). Normally, the tumor may either be small gray/red in appearance or large, polypoid, and fleshy. The botryoid variant manifests grape-like features. Histologically, they are classified into the embryonal phenotype for the vast majority of skull base rhabdomyosarcomas and are composed of primitive small round to spindled monotonous cell proliferation in sheets (Fig. 21). A myxoid stroma may be present giving rise to the botryoid variant. The alveolar phenotype is less common and is composed of tumor cells with eosinophilic cytoplasm in clusters separated by fibrous septa. Rhabdomyoblasts and multinucleated giant cells may be seen.
Differential Diagnosis Angiofibromas may be confused with vascular and inflammatory polyp, hemangiopericytomas, and vascular proliferations, hemangiomas, and pyogenic granulomas (75–77).
Myxoma These lesions are intraosseous, ill-defined lesions comprised of mucomyxoid stroma with scattered stellate cells. If a fibrous component is visible it is called myxofibroma. Although they most commonly present in the jaw, the maxilla may be affected. These lesions affect more females than males (76,77). Usually, they appear as unencapsulated, wellcircumscribed, nodular, tan to white gelatinous lesions.
Figure 20 Nasopharyngeal angiofibroma with scattered stellate cells in a fibrous stroma with prominent vessels.
52
Williams and El-Naggar Table 4 Clinical Features of Rhabdomyosarcoma Subtypes Feature
Embryonal
Site Age Outcome
Nasopharynx same sinonasal tract Children/young adults Young: 60% Adults: 10% 80%
Incidence
Alveolar Adults Poor 20%
Figure 22 Synovial sarcoma with small round primitive cells (top) and spindled cells (bottom).
Normally, the tumor presents as light tan and fleshy with a smooth surface with soft-to-firm consistency. Histologically, the tumor typically forms sweeping spindle cell interlacing bundles extending to surrounding tissue cells with moderate to low mitotic activity.
Differential Diagnosis Figure 21 Rhabdomyosarcoma, round small primitive cells forming sheets surrounding normal glands (lower left).
Ancillary Markers Immunohistochemical markers including desmin, myo-D, and myogenin are necessary for the diagnosis, especially of the embryonal form.
Differential Diagnosis Tumors that may be confused with embryonal rhabdomyosarcoma include lymphoma, neuroblastoma, Ewing sarcoma/PNET and melanoma. Immunophenotyping is critical for the diagnosis (Table 5).
Fibrosarcoma This is the most frequent mesenchymal malignancy of the sinonasal tract affecting most frequently the maxillary sinus, nasal cavity, and the ethmoid region (83–85).
A host of spindle cell benign and malignant tumors must be differentiated from this lesion including neurofibroma, rhabdomyo and synovial sarcomas, hemangiopericytoma, spindle cell carcinoma, and spindle cell melanoma. Also, reactive myofibroblastic tumors and fibromatosis should be included. Immunohistochemical markers will exclude neural, skeletal muscle, and spindle cell carcinoma if keratin is positive. Spindle cell carcinoma, however, is phenotypically more pleomorphic and may contain an epithelial component. Synovial sarcoma may exhibit mixed pattern but testing for the t(X;18) fusion gene is helpful to differentiate a purely spindle cell form.
Synovial Sarcoma Synovial sarcoma may involve the skull base region as extension from oropharynx or adjacent structures. The tumor affects the young age groups with a median of 25 years and a male predominance (86–91). Histologically, synovial sarcoma may present as a pure spindle cell variant (monophasic) or biphasic where both spindle and epithelial components are present (Fig. 22). The
Table 5 Immunohistochemical Markers useful in the Differential Diagnosis of Undifferentiated Skull Base Neoplasms Marker
Neuroblastoma
Ewing/PNET
RMS
Lymphoma
NEC
Melanoma
Keratin Synap HMB45 CD99 Desmin Myogenin S-100 CD45
–/focal +++ – – – – Focal –
– + – +++ – – – –
–/rare – – +/– +++ +++ – –
– – – – – – – +++
+ +++ ++ – – – Focal –
Rare Rare +++ – – – ++ –
Abbreviations: Ewing, Ewing sarcoma; NEC, neuroendocrine carcinoma; PNET, peripheral neuroectodermal tumor; RMS, rhabdomyosarcoma; Synap, Synaptophysin.
Chapter 2: Pathology of Tumor and Tumor-like Lesions of the Skull Base
53
Figure 23 Diagnostic Algorithm for Undifferentiated Skull Base Tumors
Table 6 Molecular Markers of Diagnostic and Therapeutic Potential in Skull Base Tumors
hybridization or polymerase chain reactions for the t(X;18) (p11.2; q11;2) translocation is complementary (Table 6).
Tumor
Alterations
Neuroblastoma 12q22-23 PNET/Ewing’ t(11;22) t(21;22) t(7;22) t(17;22) t(2;22) RMS t(2;13) (alveolar) t(1:13) Synovial Sarc t(X;18)
Gene/transcript
Incidence
Prognosis
Differential Diagnosis
hASH1 FLI-1/EWS EWS/ERG EWS/ETV1 EWS/E1AF FEV/EWS PAX3/FKHR
<90% 85% 5% Rare Rare Rare 60%
? ? ? ? ? ? Poor
This entity should be differentiated from spindle forming tumors including spindle cell carcinoma, fibrosarcoma, melanoma, and metastatic carcinoma to the base of skull. The combined features of clinicopathologic and immunohistochemical markers should be integrated in the diagnosis of this entity.
PAX7/FKHR SYT/SSX1 or SYT/SSX2
10% 64% 33%
Good Poor Poor
Abbreviations: PNET/Ewing, Peripheral neuroectodermal tumor/Ewing sarcoma; Sarc, Sarcoma
Angiosarcoma Angiosarcoma of the skull base region is rare, and if encountered, the diagnosis is based on the identification of abnormal vascular features and the use of immunohistochemical markers to support lineage (Fig. 24) (92,93).
Chordoma epithelioid component is typically formed of cuboidal or columnar epithelial cells forming cords, nests, and pseudoglandular spaces intermingled with a spindle cell proliferation.
Ancillary Markers Cytokeratins of low and high molecular weight and epithelial membrane antigen (EMA) are positive in epithelial tumor cells. In case of monomorphic spindle cells, in-situ
Chordoma is a low- to intermediate-grade malignant neoplasm originating from the notochord. The base of skull involvement is through the sphenoid occipital region, which is the site for approximately 20% of all chordomas. Lesions affecting the head and neck region frequently occur in middleage patients and those in their sixties. Patients typically present with neurological symptoms, headache, and progressive pain (94–98). Generally, chordoma is characteristically lobulated myxoid, expansive mass with mucus and a
54
Williams and El-Naggar
Figure 24 Angiosarcoma with atypical neoplastic cells forming vascular spaces.
Figure 26 Chondrosarcoma comprised of cartilage with increased cellularity of chondrocytes with a haphazard growth pattern.
slippery-appearance. Histologically, it is divided into classic, chondroid, and dedifferentiated phenotypes. The classic type is manifested by lobulated growth pattern with cords and islands of polygonal and vacuolated cells in myxomucoid background (Fig. 25). The characteristic vacuolated and eosinophilic cell is called a physaliphorous cell. The chondroid type manifests the same features along with areas of hyaline cartilaginous tissue. Dedifferentiated signifying transformation to a high-grade sarcoma shows marked cytologic atypia and high cellularity.
the reactivity to keratin and EMA is helpful in the exclusion of chondrosarcoma.
Ancillary Markers Chordoma is characteristically immunoreactive to cytokeratin, S-100, and EMA markers.
Differential Diagnosis Chordoma should be differentiated from mucinous adenocarcinoma, myxoma, and cartilaginous neoplasms. Immunohistochemical stains typically aid in the diagnosis. In particular,
Chondrosarcoma Chondrosarcoma may present as an extension from a maxillary primary. They present in a wide range of age groups, with the mesenchymal phenotype affecting mainly patients in the second and third decades of life. The most common presenting symptoms are craniofacial bone expansion and pain (99–101). Typically, these tumors manifest as translucent, cartilaginous, and scattered calcifications. Myxomatous areas with lobulation are commonly seen. The histologic spectrum seen in these tumors ranges from benign appearing hyaline cartilaginous lesions to highly cellular malignant spindle cell sarcoma. The characteristic malignant chondrocytic cells must be identified (Fig. 26). The tumor can manifest as myxoid, clear cell, dedifferentiated, and mesenchymal phenotypes. Mesenchymal chondrosarcoma is rare and may cause differential diagnostic difficulties. These tumors are composed of highly cellular spindle cell proliferations in interlacing short fascicles with focal cartilaginous formations.
Differential Diagnosis The main differential diagnosis includes enchondroma, osteochondromas, and chondroblastic osteosarcoma and chordomas. Mesenchymal chondrosarcoma should be differentiated from spindle cell malignant tumors and primitive sinonasal tumors. CD99 and SOX9 markers are typically positive in mesenchymal chondrosarcoma and could be of use in the diagnosis of this entity.
Osteosarcoma
Figure 25 Chordoma comprised of vacuolated physaliphorous cells in a myxoid background.
Osteosarcoma is a rare tumor in the skull base region. The tumor most commonly represents an extension from maxillary origin (102,103). Usually, osteosarcoma manifests as an ill-defined, irregular, and tan-yellow tissue mass with gritty (bone) sensation. Histologically, these tumors exhibit malignant cellular proliferation with osteoid bone formation (Fig. 27). The degree of cellularity and anaplastic cellular features reflect the grade of these tumors. The most common type of osteosarcoma is the osteogenic phenotype.
Chapter 2: Pathology of Tumor and Tumor-like Lesions of the Skull Base
Figure 27 Osteosarcoma is hallmarked by the presence of osteoid admixed with neoplastic cells.
55
Figure 28 Neuroblastoma comprised primitive small tumor cells.
Differential Diagnosis Differential Diagnosis This tumor should be differentiated from other bone forming lesions, including osteoblastoma, chondrosarcoma with osteoid formation, dedifferentiated chondrosarcoma, and chondroblastic osteosarcoma. Radiologic and histopathologic correlation are important.
Tumors to be differentiated from neuroblastoma include lymphoma, melanoma, small round cell (Ewing sarcoma/PNET) tumor, rhabdomyosarcoma, and adenoid cystic carcinoma (solid form). Immunohistochemical profiles including a spectrum of different cell lineage are crucial to the diagnosis (Table 5).
Ewing Sarcoma/Primitive Neuroectodermal Tumor NEUROGENIC NEOPLASMS The tumors are derived from neuroglial origin and manifest a primitive small cell growth. These morphologic similarities may lead to misclassification. The most frequently encountered entities at the skull base are neuroblastoma and the primitive neuroectodermal group of tumors.
Neuroblastoma This entity arises from neuroepithelium in the upper aspect of the nasal cavity and roof of nose and the cribriform plate of the ethmoid sinus. They comprise approximately 5% sinonasal tract malignancies. Neuroblastoma affects both genders equally with bimodal age clustering at the first and second and the fourth and fifth decades of life. They typically present as a unilateral nasal mass with obstruction and bleeding symptoms (104–109). Comparison with other primitive sinonasal tumors is shown in Table 7. Grossly, neuroblastomas are light tan and soft tissue masses. Histologically, tumors are composed of small uniform sheets and nests of primitive basal cells with minimal cytoplasm with neurofibrillary background and occasionally with neuroepithelial pseudo-resetting features (Homer– Wright structure) (Fig. 28). True rosettes formation with ductlike spaces (Flexner–Wintersteiner rosette) is rare. High-grade tumors are characterized by large pleomorphic cells and necrosis.
Ancillary Markers Negative staining for keratin, synaptophysin, and other neuroendocrine and muscle markers establish the diagnosis of neuroblastoma (Table 7). Amplification of c-Myc oncogene and loss of chromosome 1p have been considered poor prognostic markers.
Ewing sarcoma and primitive neuroectodermal tumor are inter-related primitive round cell malignancies of neuronectodermal derivation. They represent a spectrum of morphologic entities that share common molecular genetic features. They are uncommon childhood and young adult tumors affecting the skull base and the sinonasal tract regions in approximately 5% of the patients. The maxillary sinus and the nasal fossa are the common affected sites (110–115). Generally, these tumors present as light-tan, soft, and fleshy tissues with hemorrhage and mucosal ulceration. Histologically, the tumor presents in sheet and nests of densely uniform small cell proliferation (Fig. 29).
Ancillary Markers CD99 (MIC2) is generally positive; NSE and synaptophysin are less often expressed. PCR-based methods to detect the EWS/FLI gene fusion transcript and in-situ hybridization of chromosomes t(11;22) or t(21;22) are helpful in confirming the diagnosis (Table 6).
Differential Diagnosis The differential diagnosis includes all small round cells tumors, lymphoma, melanoma, rhabdomyosarcoma, small cell carcinoma, and pituitary adenoma. A combined cytomorphologic and immunohistochemical markers panel is sufficient for establishing the diagnosis (Table 5).
NEUROENDOCRINE NEOPLASMS Tumors of neuroendocrine derivation comprise a spectrum of histomorphologic entities that include typical carcinoid (well differentiated) neuroendocrine carcinoma, atypical carcinoid (moderately differentiated) neuroendocrine carcinoma and
56
Williams and El-Naggar
Figure 29 Ewing sarcoma/PNET growing in sheets of monotonous small cells.
Figure 31 Paraganglioma with nested neuroendocrine cells (Zellballen) surrounded by thin vasculature.
small cell neuroendocrine carcinoma (Table 2). These are uncommon tumors at the sinonasal and skull base sites but frequently lead to differential diagnostic difficulties (116–118).
Differential Diagnosis
Carcinoid Tumors Histologically, carcinoid tumors are composed of organoid structures, including cell nests, glandular structures, and cords of monotonous basaloid cells with clear or granular cytoplasm. Atypical carcinoid lesions also exhibit these organized features but also have evidence of mitotic activity, cellular pleomorphism, and necrosis.
Small Cell Carcinoma Small cell carcinoma is generally composed of undifferentiated small cell proliferations, lacking organization with high mitotic activities and necrosis (Fig. 30). The differential diagnosis of these tumors depends on the state of differentiations for carcinoid and atypical carcinoid (119).
The differential diagnosis may include adenocarcinoma and pituitary adenoma. For small cell neuroendocrine carcinoma, a host of small undifferentiated tumors of different lineages should be included in the diagnosis. Immunohistochemical, molecular, as well as histomorphologic characteristics should be integrated.
Paraganglioma Sinonasal and skull base paraganglioma are extremely rare. They are likely to be derived from dispersed neuroendocrine cells with the sinonasal mucosal covering. Tumors at this location may behave in an aggressive fashion (120–122). The histologic characteristics of these tumors are typical of those at traditional sites with classical Zellballen organization and vascularization (Fig. 31).
Ancillary Markers Immunohistochemically, cells comprising these tumors are positive for neuroendocrine markers and negative for keratin. A helpful feature is the positivity for S-100 protein in sustentacular cells bordering the Zellballen.
MELANOMA
Figure 30 Small cell carcinoma showing primitive cells in sheets.
Primary melanoma in the sinonasal tract accounts for less than 1% of all melanomas. They afflict patients in their fifth and sixth decades of life with equal gender distribution. The most frequently affected site is the anterior nasal septum and the maxillary antrum. Symptoms are nasal obstruction, epistaxis, and nasal mass or polyp (Tables 7 and 8) (123–127). Melanomas usually appear small to large sized polypoid mass with light-tan, brown or black colorations. Histologically, the cytomorphologic features are identical to those of melanoma of the skin where spindle, rounded, and epithelioid cells forming nests, sheets and fascicles may be found (Fig. 32). These phenotypes may or may not manifest melanin pigmentation. Mucosal involvement and epidermoid migration of melanocytic cells is a helpful diagnostic feature when present.
Chapter 2: Pathology of Tumor and Tumor-like Lesions of the Skull Base
57
Table 7 Clinicopathologic Features of Sinonasal Small Round Cell Neoplasms Feature Age (years) Site Cranial neuropathy Anaplasia Necrosis Vascular invasion
Neuroblastoma 10–20 & 50 Cribriform plate Rare Occasional Rare Rare
Ewing/PNET <30 Maxillary Rare Rare Common Rare
RMS <20 Any Not uncommon Common Rare Rare
Lymphoma 50–60 Any Common Common Common Common
Melanoma 45–65 Any Rare Common Rare Common
Abbreviations: Ewing: Ewing sarcoma; PNET, peripheral neuroectodermal tumor; RMS, Rhabdomyosarcoma. Table 8 Characteristics of Mucosal Melanoma •<1% of all melanomas •<5% of Sinonasal tumors •M/F: 1:1 •Age: Fifth to eighth decades •Race: Japanese •Site: •Anterior septum •Maxillary antrum •Poor prognosis •Adverse features: •>3.0 cm size •Advanced age •Vascular invasion
LYMPHOPROLIFERATIVE DISORDERS Sinus Histiocytosis with Massive Lymphadenopathy (Rosai–Dorfman Disease) This is a rare, idiopathic, histiocytic, proliferative disorder affecting typically young black women. This disease affects the paranasal sinuses, the nasal cavity, and the eyes. Patients present with nasal obstruction, proptosis, cranial nerve deficits, mass, and fever (128–130). Usually, lymphoreticular tumors appear as polypoid, gray to yellow soft tissue masses. Histologically, they show marked sinus expansion with lymphoplasmacytic proliferation. The sinuses are also filled with histiocytic cell with lymphocytes and red cells in their cytoplasm (lymphophagocytosis/emperipolesis).
Ancillary Markers Ancillary Markers This tumor is negative for keratin and reactive to melanocytic markers including HMB-45, Melan-A, MART-1, tyrosinase, and S-100.
Differential Diagnosis Melanoma at these locations should be differentiated from metastatic melanoma and primary undifferentiated skull base neoplasms including undifferentiated carcinoma, neuroendocrine tumor, neuroblastoma, lymphoma, and small peripheral neuroectodermal tumors and rhabdomyosarcoma. Immunohistochemical markers are fundamental in differentiating these tumors (Table 5).
The diagnosis may be aided by staining of histiocytic markers including S-100 protein and CD68.
Differential Diagnosis Rosai–Dorfman disease must be differentiated from infectious entities, such as rhinoscleroma, lepromatous leprosy, and malignant, histiocytic lymphoma.
Lymphoma Although rare, most lymphoproliferative lesion arise from the skull base region (131–133) are B-cell lineage lymphomas and sinonasal NK/T-cell lymphoma. These tumors may affect the nasal cavity, paranasal sinuses and may extend to involve any adjacent structures. Patients present with nasal obstruction, epistaxis, proptosis, or mass lesion. NK/T-cell lymphoma is seen in adults with a male predominance. Lymphoproliferative tumors appear grossly as soft, tan-gray and often described as “fish-flesh” soft tissue masses. Histology will vary based on the underlying entity, however, most will show sheets of atypical lymphoid cells with a diffuse growth pattern infiltrating the stroma (Fig. 33). The tumor cells are frequently angiocentric around blood vessels in the NK/T-cell type. Prominent necrosis is also common in this specific entity, leading to a midline facial destructive process.
Ancillary Markers
Figure 32 Melanoma showing prominent nucleoli and scattered pigment (melanin) growing in sheets.
The diagnosis is aided by staining of lymphoid markers including CD45 and lineage markers (i.e., B-cell CD20; T-cell CD3, CD4, and CD8) among a multitude of other markers like CD56. On histologic samples, EBV can be identified by in-situ hybridization for EBV-encoded RNA, which is almost universally positive in the NK/T-cell type lymphoma and may be associated with some diffuse B-cell lymphomas. When lymphoma is suspected, fresh tissue should be sent for flow-cytometric analysis along with tissue for permanent histologic examination.
58
Williams and El-Naggar
Figure 33 Diffused B-cell lymphoma composed of sheets of pleomorphic cells and necrosis.
Differential Diagnosis This lesion may mimic small, round, blue cell tumors (Table 7) and undifferentiated carcinomas (sinonasal and nasopharyngeal). If prominent necrosis is present, Wegener granulomatosis and other chronic lesion may be considered. IHC for carcinoma and small round cell tumor and fungal stains are excluding these entities.
MISCELLANEOUS Teratocarcinosarcoma This is a rare skull base and sinonasal tract malignant tumor composed of carcinomatous, sarcomatous, and immature neural elements. The most frequent sites are the ethmoid and maxillary sinuses and the nasal cavity in elderly male patients. Grossly, these rare tumors appear bulky, polypoid, and friable light-tan to red mass. Histologically, the tumor is characteristically composed of a high-grade carcinomatous component admixed with sarcomatous and immature neural elements (Fig. 34). Benign and malignant germ cell elements may also be present. The carcinoma can be either squamous or adeno or neuroendocrine, and the sarcoma can be cartilaginous, bony or skeletal muscle in nature. The neural elements may contain primitive rosettes with neurofibrillary features (134,135).
Figure 34 Teratocarcinosarcoma showing mixed high-grade neoplasms with nests of carcinoma and spindled sarcoma cells between the nests.
Ameloblastoma Ameloblastoma is a tumor of intermediate malignancy and odontogenic origin. They may arise from enamel organ, residual dental lamina, epithelial lining of dentigerous or follicular cysts, or rarely heterotopic embryonal enamel in the sinonasal tract. Ameloblastoma at the base of skull are typically an extension from tumors originating in the molar region of the maxillary antrum. They may frequently present with nasal symptoms without history of dental complaints (138,139). Histologically, the solid form manifests cellular, follicular structures with palisading basal cells with branding cords and cellular islets (Fig. 36). The cystic form manifests varied spaces lined with palisading basal cells with squamous metaplasia with central satellite appearing cells. Extraosseous ameloblastoma may histologically mimic basal cell carcinoma, and efforts to exclude this possibility must be made.
Differential Diagnosis This tumor causes a differential diagnostic dilemma if a dominant component is present on small biopsy samples. The diagnosis may depend on the available tissue component.
Meningioma Histologically, skull base meningiomas are identical morphologically to their primary meningeal origin but tend to be locally invasive, frequently showing characteristic whirling (Fig. 35). They are typically meningothelial or fibroblastic but other types may also be found (136,137).
Figure 35 Meningioma forming nests of small uniform cells within the stroma and vascular spaces.
Chapter 2: Pathology of Tumor and Tumor-like Lesions of the Skull Base
Figure 36 Ameloblastoma showing the palisading basaloid epithelial proliferation with central stellate reticulum mimicking dental elements.
METASTATIC NEOPLASMS Isolated metastatic tumor to this region is very rare. It may, however, occur among metastasis to multiple other sites. Renal cell carcinoma is by far the most common source. Others including breast, lung, melanoma, and testicular tumors follow renal in their frequency. Isolated incidences of metastasis from various other tumors have also been reported. Metastatic tumors affect both genders equally notwithstanding gender specific influence. In female patients, breast, gynecologic, and thyroid tumors are the most frequent primary origins. In male patients, lung, prostate, kidney and bone in decreasing order are the most common sites (140–142). REFERENCES 1. Lai SY, Kennedy, DW, Bolger, WE. Sphenoid encephaloceles: Disease management and identification of lesions within the lateral recess of the sphenoid sinus. Laryngoscope. 2002;112(10): 1800–1885. 2. Jabre A, Tabaddor R, Samaraweera R. Transsphenoidal meningoencephalocele in adults. Surg Neurol. 2000;54(2):183–187. Discussion 187–188. 3. Penner CR, Thompson LD. Nasal glial heterotopia: A clinicopathologic and immunophenotypic analysis of 10 cases with a review of the literature. Ann Diagn Pathol. 2003;7(6):354–359. 4. Mortuaire G, Pasquesoone X, Leroy X, et al. Respiratory epithelial adenomatoid hamartomas of the sinonasal tract. Eur Arch Otorhinolaryngol. 2007;264(4):451–453. 5. Ozolek JA, Hunt JL. Tumor suppressor gene alterations in respiratory epithelial adenomatoid hamartoma (REAH): Comparison to sinonasal adenocarcinoma and inflamed sinonasal mucosa. Am J Surg Pathol. 2006;30(12):1576–1580. 6. Graeme-Cook F, Pilch BZ. Hamartomas of the nose and nasopharynx. Head Neck. 1992;14(4):321–327. 7. Pasquini E, Faustini-Fustini M, Sciarretta V, et al. TSH-secreting pituitary adenoma of the vomerosphenoidal junction. Eur J Endocrinol. 2003;148(2):253–257. 8. Lloyd RV, Chandler WF, Kovacs K, et al. Ectopic pituitary adenomas with normal anterior pituitary glands. Am J Surg Pathol. 1986;10(8):546–552. 9. McCall T, Fassett DR, Lyons G, et al. Inflammatory pseudotumor of the cavernous sinus and skull base. Neurosurg Rev. 2006; 29(3):194–200.
59
10. Newlin HE, Werning JW, Mendenhall WM. Plasma cell granuloma of the maxillary sinus: A case report and literature review. Head Neck. 2005;27(8):722–728. 11. Maruya S, Kurotaki H, Hashimoto T, et al. Inflammatory pseudotumour (plasma cell granuloma) arising in the maxillary sinus. Acta Otolaryngol. 2005;125(3):322–327. 12. Polzehl D, Moeller P, Riechelmann H, et al. Distinct features of chronic rhinosinusitis with and without nasal polyps. Allergy. 2006;61(11):1275–1279. 13. Garavello W, Gaini RM. Histopathology of routine nasal polypectomy specimens: A review of 2,147 cases. Laryngoscope. 2005;115(10):1866–1868. 14. Batsakis JG, Sneige N. Choanal and angiomatous polyps of the sinonasal tract. Ann Otol Rhinol Laryngol. 1992;101(7):623–625. 15. Lawson W, Kaufman MR, Biller HF. Treatment outcomes in the management of inverted papilloma: An analysis of 160 cases. Laryngoscope. 2003;113(9):1548–1556. 16. Kaufman MR, Brandwein MS, Lawson W. Sinonasal papillomas: Clinicopathologic review of 40 patients with inverted and oncocytic Schneiderian papillomas. Laryngoscope. 2002;112(8PT1):1372–1377. 17. Kraft M, Simmen D, Casas R, et al. Significance of human papillomavirus in sinonasal papillomas. J Laryngol Otol. 2001; 115(9):709–714. 18. Califano J, Koch W, Sidransky D, et al. Inverted sinonasal papilloma: A molecular genetic appraisal of its putative status as a precursor to squamous cell carcinoma. Am J Pathol. 2000; 156(1):333–337. 19. Michaels L, Young M. Histogenesis of papillomas of the nose and paranasal sinuses. Arch Pathol Lab Med. 1995;119(9):821– 826. 20. Vrabec DP. The inverted Schneiderian papilloma: A 25-year study. Laryngoscope. 1994;104(5Pt1):582–605. 21. Hartwick RW, Batsakis JG. Sinus aspergillosis and allergic fungal sinusitis. Ann Otol Rhinol Laryngol. 1991;100(5pt1):427–430. 22. Granville L, Chirala M, Cernoch P, et al. Fungal sinusitis: Histologic spectrum and correlation with culture. Hum Pathol. 2004;35(4):474–481. 23. Taxy JB. Paranasal fungal sinusitis: Contributions of histopathology to diagnosis: A report of 60 cases and literature review. Am J Surg Pathol. 2006;30(6):713–720. 24. Kontoyiannis DP, Lionakis MS, Lewis RE, et al. Zygomycosis in a tertiary-care cancer center in the era of Aspergillus-active antifungal therapy: A case-control observational study of 27 recent cases. J Infect Dis. 2005;191(8):1350–1360. 25. Parikh SL, Venkatraman G, DelGaudio JM. Invasive fungal sinusitis: A 15-year review from a single institution. Am J Rhinol. 2004;18(2):75–81. 26. Botelho-Nevers E, Gouriet F, Lepidi H, et al. Chronic nasal infection caused by Klebsiella rhinoscleromatis or Klebsiella ozaenae: Two forgotten infectious diseases. Int J Infect Dis. 2007;11(5):423– 429. 27. Quevedo J. Scleroma in Guatemala, with a study of the disease based on the experience of 108 cases. Ann Otol Rhinol Laryngol. 1949;58(3):613–645. 28. Keni SP, Wiley EL, Dutra JC, et al. Skull base Wegener’s granulomatosis resulting in multiple cranial neuropathies. Am J Otolaryngol. 2005;26(2):146–149. 29. Heffner DK. Wegener’s granulomatosis is not a granulomatous disease. Ann Diagn Pathol. 2002;6(5):329–333. 30. Bajema IM, Hagen EC, Van Der Woude FJ, et al. Wegener’s granulomatosis: A meta-analysis of 349 literary case reports. J Lab Clin Med. 1997;129(1):17–22. 31. Hoffman GS, Kerr GS, Levitt RY, et al. Wegener granulomatosis: An analysis of 158 patients. Ann Intern Med. 1992;116(6):488– 498. 32. Buchmann L, Emami A, Wei JL. Primary head and neck Langerhans cell histiocytosis in children. Otolaryngol Head Neck Surg. 2006;135(2):312–317. 33. Krishna H, Behari S, Pal L, et al. Solitary Langerhans-cell histiocytosis of the clivus and sphenoid sinus with parasellar and petrous extensions: Case report and a review of literature. Surg Neurol. 2004;62(5):447–454.
60
Williams and El-Naggar
34. Davis SE, Rice DH. Langerhans’ cell histiocytosis: Current trends and the role of the head and neck surgeon. Ear Nose Throat J. 2004;83(5):340, 342, 344. 35. Boston M, Derkay CS. Langerhans’ cell histiocytosis of the temporal bone and skull base. Am J Otolaryngol. 2002;23(4):246– 248. 36. Brisman JL, Feldstein NA, Tarbell NJ, et al. Eosinophilic granuloma of the clivus: Case report, follow-up of two previously reported cases, and review of the literature on cranial base eosinophilic granuloma. Neurosurgery. 1997;41(1):273–278. 37. Sindwani R, Cohen JT, Pilch BZ, et al. Myospherulosis following sinus surgery: Pathological curiosity or important clinical entity? Laryngoscope. 2003;113(7):1123–1127. 38. Tsai EC, Santoreneos S, Rutka JT. Tumors of the skull base in children: Review of tumor types and management strategies. Neurosurg Focus. 2002;12(5):e1. 39. Richardson MS. Pathology of skull base tumors. Otolaryngol Clin North Am. 2001;34(6):1025–1042. 40. Porceddu S, Martin J, Shanker G, et al. Paranasal sinus tumors: Peter MacCallum Cancer Institute experience. Head Neck. 2004;26(4):322–330. 41. Dulguerov P, Jacobsen MS, Allal AS, et al. Nasal and paranasal sinus carcinoma: Are we making progress? A series of 220 patients and a systematic review. Cancer. 2001;92(12):3012–3029. 42. Wieneke JA, Thompson LD, Wenig BM. Basaloid squamous cell carcinoma of the sinonasal tract. Cancer. 1999;85(4):841–854. 43. Lund VJ, Howard DJ, Wei WI, et al. Craniofacial resection for tumors of the nasal cavity and paranasal sinuses—A 17-year experience. Head Neck. 1998;20(2):97–105. 44. Frierson HF Jr. Sinonasal undifferentiated carcinoma. In: Barnes EL, Eveson JW, Reichart P, Sidransky D, eds. Pathology and Genetic of Head and Neck Tumors. Lyon, France: IARC Press, 2005:19. Kleihues P, Sobin LH, series eds. World Health Organization Classifications of Tumors. 45. Ejaz A, Wenig BM. Sinonasal undifferentiated carcinoma: Clinical and pathologic features and a discussion on classification, cellular differentiation, and differential diagnosis. Adv Anat Pathol. 2005;12(3):134–143. 46. Jeng YM, Sung MT, Fang CL, et al. Sinonasal undifferentiated carcinoma and nasopharyngeal-type undifferentiated carcinoma: Two clinically, biologically, and histopathologically distinct entities. Am J Surg Pathol. 2002;26(3):371–376. 47. Franchi A, Moroni M, Massi D, et al. Sinonasal undifferentiated carcinoma, nasopharyngeal-type undifferentiated carcinoma, and keratinizing and nonkeratinizing squamous cell carcinoma express different cytokeratin patterns. Am J Surg Pathol. 2002;26(12):1597–1604. 48. Cerilli LA, Holst VA, Brandwein MS, et al. Sinonasal undifferentiated carcinoma: Immunohistochemical profile and lack of EBV association. Am J Surg Pathol. 2001;25(2):156–163. 49. Smith SR, Som P, Fahmy A, et al. A clinicopathological study of sinonasal neuroendocrine carcinoma and sinonasal undifferentiated carcinoma. Laryngoscope. 2000;110(10Pt1):1617–1622. 50. Sckolnick J, Murphy J, Hunt JL. Microsatellite instability in nasopharyngeal and lymphoepithelial carcinomas of the head and neck. Am J Surg Pathol. 2006;30(10):1250–1253. 51. Chan JKC, Bray F, McCarron P, et al. Nasopharyngeal carcinoma. In: Barnes El-Eveson JW, Reichart P, Sidransky D, eds. Pathology and Genetics of Head and Neck Tumours. Lyon, France: IARC Press, 2005 :85–97. Kleihues P, Sobin LH, Series eds. World Health Organization Classification of Tumours. 52. Wei WI, Sham JS. Nasopharyngeal carcinoma. Lancet. 2005;365(9476):2041–2054. 53. Lo KW, To KF, Huang DP. Focus on nasopharyngeal carcinoma. Cancer Cell. 2004;5(5):423–428. 54. Shi W, Pataki I, MacMillan C, et al. Molecular pathology parameters in human nasopharyngeal carcinoma. Cancer. 2002;94(7):1997–2006. 55. Choi PH, Suen MW, Huang DP, et al. Nasopharyngeal carcinoma: Genetic changes, Epstein-Barr virus infection, or both. A clinical and molecular study of 36 patients. Cancer. 1993;72(10):2873–2878.
56. Hsu HC, Chen CL, Hsu MM, et al. Pathology of nasopharyngeal carcinoma. Proposal of a new histologic classification correlated with prognosis. Cancer. 1987;59(5):945–951. 57. Neto AG, Pineda-Daboin K, Spencer ML, et al. Sinonasal acinic cell carcinoma: A clinicopathologic study of four cases. Head Neck. 2005;27(7):603–607. 58. Hampton TA, Scheithauer BW, Rojiani AM, et al. Salivary gland-like tumors of the sellar region. Am J Surg Pathol. 1997;21(4):424–434. 59. Yom SS, Rashid A, Rosenthal DI, et al. Genetic analysis of sinonasal adenocarcinoma phenotypes: Distinct alterations of histogenetic significance. Mod Pathol. 2005;18(3):315–319. 60. Cathro HP, Mills SE. Immunophenotypic differences between intestinal-type and low-grade papillary sinonasal adenocarcinomas: An immunohistochemical study of 22 cases utilizing CDX2 and MUC2. Am J Surg Pathol. 2004;28(8):1026–1032. 61. Choi HR, Sturgis EM, Rashid A, et al. Sinonasal adenocarcinoma: Evidence for histogenetic divergence of the enteric and nonenteric phenotypes. Hum Pathol. 2003;34(11):1101–1107. 62. Neto AG, Pineda-Daboin K, Luna MA. Sinonasal tract seromucous adenocarcinomas: A report of 12 cases. Ann Diagn Pathol. 2003;7(3):154–159. 63. Franchi A, Gallo O, Santucci M. Clinical relevance of the histological classification of sinonasal intestinal-type adenocarcinomas. Hum Pathol. 1999;30(10):1140–1145. 64. Franquemont DW, Fechner RE, Mills SE. Histologic classification of sinonasal intestinal-type adenocarcinoma. Am J Surg Pathol. 1991;15(4):368–375. 65. Miller FR, D’Agostino MA, Schlack K. Lobular capillary hemangioma of the nasal cavity. Otolaryngol Head Neck Surg. 1999; 120(5):783–784. 66. Mills SE, Cooper PH, Fechner RE. Lobular capillary hemangioma: The underlying lesion of pyogenic granuloma. A study of 73 cases from the oral and nasal mucous membranes. Am J Surg Pathol. 1980;4(5):470–479. 67. Thompson LD, Miettinen M, Wenig BM. Sinonasal-type hemangiopericytoma: A clinicopathologic and immunophenotypic analysis of 104 cases showing perivascular myoid differentiation. Am J Surg Pathol. 2003;27(6):737–749. 68. Tse LL, Chan JK. Sinonasal haemangiopericytoma-like tumour: A sinonasal glomus tumour or a haemangiopericytoma? Histopathology. 2002;40(6):510–517. 69. Osammor JY, Howat AJ. Nasal haemangiopericytoma. J Laryngol Otol. 1991;105(7):593–595. 70. Batsakis JG, Jacobs JB, Templeton AC. Hemangiopericytoma of the nasal cavity: Electron-optic study and clinical correlations. J Laryngol Otol. 1983;97(4):361–368. 71. Brunnemann RB, Ro JY, Ordonez NG, et al. Extrapleural solitary fibrous tumor: A clinicopathologic study of 24 cases. Mod Pathol. 1999;12(11):1034–1042. 72. Zukerberg LR, Rosenberg AE, Randolph G, et al. Solitary fibrous tumor of the nasal cavity and paranasal sinuses. Am J Surg Pathol. 1991;15(2):126–130. 73. Abraham SC, Montgomery EA, Giardiello FM, et al. Frequent beta-catenin mutations in juvenile nasopharyngeal angiofibromas. Am J Pathol. 2001;158(3):1073–1078. 74. Beham A, Beham-Schmid C, Regauer S, et al. Nasopharyngeal angiofibroma: True neoplasm or vascular malformation? Adv Anat Pathol. 2000;7(1):36–46. 75. Hwang HC, Mills SE, Patterson K, et al. Expression of androgen receptors in nasopharyngeal angiofibroma: An immunohistochemical study of 24 cases. Mod Pathol. 1998;11(11):1122– 1126. 76. Sato H, Gyo K, Tomidokoro Y, et al. Myxoma of the sphenoidal sinus. Otolaryngol Head Neck Surg. 2004;130(3):378–380. 77. Andrews T, Kountakis SE, Maillard AA. Myxomas of the head and neck. Am J Otolaryngol. 2000;21(3):184–189. 78. Lezzoni JC, Mills SE. “Undifferentiated” small round cell tumors of the sinonasal tract: Differential diagnosis update. Am J Clin Pathol. 2005;124(Suppl):S110–S121. 79. Sorensen PH, Lynch JC, Qualman SJ, et al. PAX3-FKHR and PAX7-FKHR gene fusions are prognostic indicators in
Chapter 2: Pathology of Tumor and Tumor-like Lesions of the Skull Base
80.
81.
82.
83.
84.
85.
86.
87. 88. 89.
90.
91.
92.
93.
94.
95.
96.
97.
98. 99.
100.
101.
102.
alveolar rhabdomyosarcoma: A report from the children’s oncology group. J Clin Oncol. 2002;20(11):2672–2679. Parham DM. Pathologic classification of rhabdomyosarcomas and correlations with molecular studies. Mod Pathol. 2001;14(5):506–514. Callender TA, Weber RS, Janjan N, et al. Rhabdomyosarcoma of the nose and paranasal sinuses in adults and children. Otolaryngol Head Neck Surg. 1995;112(2):252–257. El-Naggar AK, Batsakis JG, Ordonez NG, et al. Rhabdomyosarcoma of the adult head and neck: A clinicopathological and DNA ploidy study. J Laryngol Otol. 1993;107(8):716–720. Gnepp DR, Henley J, Weiss S, et al. Desmoid fibromatosis of the sinonasal tract and nasopharynx. A clinicopathologic study of 25 cases. Cancer. 1996;78(12):2572–2579. Heffner DK, Gnepp DR. Sinonasal fibrosarcomas, malignant schwannomas, and “Triton” tumors. A clinicopathologic study of 67 cases. Cancer. 1992;70(5):1089–1101. Frankenthaler R, Ayala AG, Hartwick RW, et al. Fibrosarcoma of the head and neck. Laryngoscope. 1990;100(8):799– 802. Simunjak B, Petric V, Bedekovic V, et al. Dimensions and outcome of synovial sarcoma of the head and neck: Case presentation and review of the literature. J Otolaryngol. 2005;34(6):420– 423. Gallia GL, Sciubba DM, Hann CL, et al. Synovial sarcoma of the frontal sinus. Case report. J Neurosurg. 2005;103(6):1077–1080. Potter BO, Sturgis EM. Sarcomas of the head and neck. Surg Oncol Clin N Am. 2003;12(2):379–417. Argani P, Zakowski MF, Klimstra DS, et al. Detection of the SYTSSX chimeric RNA of synovial sarcoma in paraffin-embedded tissue and its application in problematic cases. Mod Pathol. 1998;11(1):65–71. Hirsch RJ, Yousem DM, Loevner LA, et al. Synovial sarcomas of the head and neck: MR findings. Am J Roentgenol. 1997;169(4):1185–1188. Cihak RA, Lydiatt WM, Lydiatt DD, et al. Synovial sarcoma of the head and neck: Chromosomal translation (X;18) as a diagnostic aid. Head Neck. 1997;19(6):549–553. Wong KF, So CC, Wong N, et al. Sinonasal angiosarcoma with marrow involvement at presentation mimicking malignant lymphoma: Cytogenetic analysis using multiple techniques. Cancer Genet Cytogenet. 2001;129(1):64–68. Velegrakis GA, Panayiotides JG, Skoulakis CE, et al. Angiosarcoma of the maxillary sinus. J Laryngol Otol. 2000;114(5):381– 384. Hoch BL, Nielsen GP, Liebsch NJ, et al. Base of skull chordomas in children and adolescents: A clinicopathologic study of 73 cases. Am J Surg Pathol. 2006;30(7):811–818. St Martin M, Levine SC. Chordomas of the skull base: Manifestations and management. Curr Opin Otolaryngol Head Neck Surg. 2003;11(5):324–327. Hug EB. Review of skull base chordomas: Prognostic factors and long-term results of proton-beam radiotherapy. Neurosurg Focus. 2001;10(3):E11. DeMonte F, Diaz E Jr, Callender D, Suk I. Transmandibular, circumglossal, retropharyngeal approach for chordomas of the clivus and upper cervical spine. Technical note. Neurosurg Focus. 2001;10(3):E10. Colli BO, Al-Mefty O. Chordomas of the skull base: Follow-up review and prognostic factors. Neurosurg Focus. 2001;10(3):E1. Sekhar LN, Pranatartiharan R, Chanda A, et al. Chordomas and chondrosarcomas of the skull base: Results and complications of surgical management. Neurosurg Focus. 2001;10(3):E2. Korten AG, ter Berg HJ, Spincemaille GH, et al. Intracranial chondrosarcoma: Review of the literature and report of 15 cases. J Neurol Neurosurg Psychiatry. 1998;65(1):88–92. Rapidis AD, Archondakis G, Anteriotis D, et al. Chondrosarcomas of the skull base: Review of the literature and report of two cases. J Craniomaxillofac Surg. 1997;25(6):322–327. Panda NK, Jain A, Eshwara R eddy CE. Osteosarcoma and chondrosarcoma of the maxilla. Br J Oral Maxillofac Surg. 2003;41(5):329–333.
61
103. Whitehead RE, Melhem ER, Kasznica J, et al. Telangiectatic osteosarcoma of the skull base. Am J Neuroradiol. 1998;19(4):754– 757. 104. Wenig BM, Dulgueron P, Kapadia SB, et al. Neuroendocrine tumours. In: Barnes El-Eveson JW, Reichart P, Sidransky D, eds. World Health Organization Classification of Tumours. Lyon, France: IARC Press, 2005 :65–75. 105. Diaz EM Jr, Johnigan RH III, Pero C, et al. Olfactory neuroblastoma: The 22-year experience at one comprehensive cancer center. Head Neck. 2005;27(2):138–149. 106. Hirose T, Scheithauer BW, Lopes MB, et al. Olfactory neuroblastoma. An immunohistochemical, ultrastructural, and flow cytometric study. Cancer. 1995;76(1):4–19. 107. Frierson HF Jr, Ross GW, Mills SE, et al. Olfactory neuroblastoma. Additional immunohistochemical characterization. Am J Clin Pathol. 1990;94(5):547–553. 108. O’Connor TA, McLean P, Jullard GJF, et al. Olfactory neuroblastoma. Cancer. 1989;63(12):2426–2428. 109. Mills SE, Frierson HF Jr. Olfactory neuroblastoma. A clinicopathologic study of 21 cases. Am J Surg Pathol. 1985;9(5):317– 327. 110. Mhawech-Fauceglia P, Herrmann F, Penetrante R, et al. Diagnostic utility of FLI-1 monoclonal antibody and dual-colour, breakapart probe fluorescence in situ (FISH) analysis in Ewing’s sarcoma/primitive neuroectodermal tumour (EWS/PNET). A comparative study with CD99 and FLI-1 polyclonal antibodies. Histopathology. 2006;49(6):569–575. 111. Windfuhr JP. Primitive neuroectodermal tumor of the head and neck: Incidence, diagnosis, and management. Ann Otol Rhinol Laryngol. 2004;113(7):533–543. 112. Csokonai LV, Liktor B, Arato G, et al. Ewing’s sarcoma in the nasal cavity. Otolaryngol Head Neck Surg. 2001;125(6):665–667. 113. Dagher R, Pham TA, Sorbara L, et al. Molecular confirmation of Ewing sarcoma. J Pediatr Hematol Oncol. 2001;23(4);221–224. 114. O’Sullivan MJ, Perlman EJ, Furman J, et al. Visceral primitive peripheral neuroectodermal tumors: A clinicopathologic and molecular study. Hum Pathol. 2001;32(10):1109–1115. 115. Toda T, Atari E, Sadi AM, et al. Primitive neuroectodermal tumor in sinonasal region. Auris Nasus Larynx. 1999;26(1):83– 90. 116. Barnes L. Neuroendocrine tumours. In: Barnes El-Eveson JW, Reichart P, Sidransky D, eds. Pathology and Genetics of Head and Neck Tumors. Lyon, France: IARC Press, 2005 :125–139. Kleihues P, Sobin LH, series eds. World Health Organization Classification of Tumours. 117. Mills SE. Neuroectodermal neoplasms of the head and neck with emphasis on neuroendocrine carcinomas. Mod Pathol. 2002; 15(3):264–278. 118. Fitzek MM, Thornton AF, Varvares M, et al. Neuroendocrine tumors of the sinonasal tract. Results of a prospective study incorporating chemotherapy, surgery, and combined proton-photon radiotherapy. Cancer. 2002;94(10):2623–2634. 119. Perez-Ordonez B, Caruana SM, Huvos AG, et al. Small cell neuroendocrine carcinoma of the nasal cavity and paranasal sinuses. Hum Pathol. 1998;29(8):826–832. 120. Balatsouras DG, Eliopoulos PN, Economou CN. Multiple glomus tumours. J Laryngol Otol. 1992;106(6):538–543. 121. Woodruff JM, Huvos AG, Erlandson RA, et al. Neuroendocrine carcinomas of the larynx. A study of two types, one of which mimics thyroid medullary carcinoma. Am J Surg Pathol. 1985; 9(11):771–790. 122. Lack EE, Cubilla AL, Woodruff JM. Paragangliomas of the head and neck region. A pathologic study of tumors from 71 patients. Hum Pathol. 1979;10(2):191–218. 123. Thompson LD, Wieneke JA, Miettinen M. Sinonasal tract and nasopharyngeal melanomas: A clinicopathologic study of 115 cases with a proposed staging system. Am J Surg Pathol. 2003;27(5):594–611. 124. Prasad ML, Jungbluth AA, Iversen K, et al. Expression of melanocytic differentiation markers in malignant melanomas of the oral and sinonasal mucosa. Am J Surg Pathol. 2001;25(6):782–787.
62
Williams and El-Naggar
125. Batsakis JG, Suarez P, El-Naggar AK. Mucosal melanomas of the head and neck. Ann Otol Rhinol Laryngol. 1998;107(7):626–630. 126. Billings KR, Wang MB, Sercarz JA, et al. Clinical and pathologic distinction between primary and metastatic mucosal melanoma of the head and neck. Otolaryngol Head Neck Surg. 1995;112(6):700–706. 127. Franquemont DW, Mills SE. Sinonasal malignant melanoma. A clinicopathologic and immunohistochemical study of 14 cases. Am J Clin Pathol. 1991;96(6):689–697. 128. Yoon AJ, Parisien M, Feldman F, et al. Extranodal RosaiDorfman disease of bone, subcutaneous tissue and paranasal sinus mucosa with a review of its pathogenesis. Skeletal Radiol. 2005;34(10):653–657. 129. Kademani D, Patel SG, Prasad ML, et al. Intraoral presentation of Rosai-Dorfman disease: A case report and review of the literature. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2002;93(6):699–704. 130. Shemen L, D’Anton M, Klijian A, et al. Rosai-Dorfman disease involving the premaxilla. Ann Otol Rhinol Laryngol. 1991;100(10):845–851. 131. Rodriguez J, Romaguera JE, Manning J, et al. Nasal-type T/NK lymphomas: A clinicopathologic study of 13 cases. Leuk Lymphoma. 2000;39(1–2):139–144. 132. Jaffe ES, Chan JK, Su IJ, et al. Report of the Workshop on Nasal and Related Extranodal Angiocentric T/Natural Killer Cell Lymphomas. Definitions, differential diagnosis, and epidemiology. Am J Surg Pathol. 1996;20(1):103–111. 133. Abbondanzo SL, Wenig BM. Non-Hodgkin’s lymphoma of the sinonasal tract. A clinicopathologic and immunophenotypic study of 120 cases. Cancer. 1995;75(6):1281–1291.
134. Shimazaki H, Aida S, Tamai S, et al. Sinonasal teratocarcinosarcoma: Ultrastructural and immunohistochemical evidence of neuroectodermal origin. Ultrastruct Pathol. 2000;24(2):115– 122. 135. Pai SA, Naresh KN, Masih K, et al. Teratocarcinosarcoma of the paranasal sinuses: A clinicopathologic and immunohistochemical study. Hum Pathol. 1998;29(7):718– 722. 136. Kjeldsberg CR, Minckler J. Meningiomas presenting as nasal polyps. Cancer. 1972;29(6):153–156. 137. Ho KL. Primary meningioma of the nasal cavity and paranasal sinuses. Cancer. 1980;46(6):1442–1447. 138. Gardner DG, Heikinheima K, Shear M, et al. Ameloblastomas. In: Barnes El-Eveson JW, Reichart P, Sidransky D, eds. Pathology and Genetics of Head and Neck Tumors. Lyon, France: IARC Press, 2005 :296–300. Kleihues P, Sobin LH, series eds. World Health Organization Classification of Tumours. 139. Schafer DR, Thompson LD, Smith BC, et al. Primary ameloblastoma of the sinonasal tract: A clinicopathologic study of 24 cases. Cancer. 1998;82(4):667–674. 140. Neville BW, Dann DD, Allen CM, et al. eds. Oral and Maxillofacial Pathology. Bethesda: Armed Forces Institute of Pathology, 2001: 263–265. Rosai J, Sobin LH, Series ed. Third series, Fascide 29. 141. Kent SE, Majumdar B. Metastatic tumours in the maxillary sinus. A report of two cases and a review of the literature. J Laryngol Otol. 1985;99(5):459–462. 142. Flocks RH, Boatman DC. Incidence of head and neck metastasis from genito-urinary neoplasms. Laryngoscope. 1973;83(9):1527–1539.
3 Genetic Abnormalities of Skull Base Tumors Ziv Gil and Dan M. Fliss
requires fresh tumor samples that can be preserved in a proper medium [i.e., Roswell Park Memorial Institute (RPMI) 1640 with 10–20% fetal calf serum] for 24–36 hours. The samples may be kept on ice, but not frozen, since freezing of the tissue inhibits its ability to grow in culture. Following tissue removal at surgery, the tumor sample is mechanically disintegrated and digested in collagenase, plated and incubated at 37◦ C in 5% CO2 for 6 to 10 days. A chromosome-banding configuration consists of alternating light and dark lines or bands that appear along its length after being stained with a dye. Each chromosome is identified by a unique banding pattern (Fig. 1). Chromosomal aberrations, including chromosome breakage, loss, duplication, or inversions, can be identified according to the representative chromosomal segments revealed by the staining technique. Giemsa has become the most popular staining technique for staining of chromosomes. Staining a metaphase chromosome with a Giemsa stain is referred to as “G-banding.” G-banding preferentially stains the regions of DNA that are rich in adenine and thymine, have high affinity for the dye, and appear as dark bands under light microscope (Fig. 2). The regions of the chromosome that are rich in guanine and cytosine have little affinity for the dye and appear as pallid bands. A standard G-band staining technique allows between 400 and 600 bands to be seen on metaphase chromosomes. With high-resolution G-banding techniques, as many as 2000 different bands have been demonstrated on human chromosomes. The main advantage of the G-banding technique is its simplicity and robustness, which gives the investigator precise resolution for the chromosomal deviations often found in human neoplasms.
John Cruso defined cancer in his book Easy Medicine (published in 1699 by the Royal College of Physicians in London) as: “A malignant ulcer, which the surgeons pronounced a cancer, but could not cure it.” Nowadays, cancer is considered a genetic disease that is characterized by deviation from normal of one or more gene products responsible for the regulation of cell growth and development. Some types of cancer are considered polygenic and multifactorial, originating from an accumulation of molecular changes in both the tumor and the host cells, leading to genetic instability and carcinogenesis. Tumor development is manifested in many cases as chromosomal changes that can be detected microscopically or with the aid of molecular techniques. Karyotypic changes involved in neoplastic transformation include numerical or structural chromosomal changes that tend to progress during time in certain tumors and contribute to malignant transformation. Chromosomal abnormalities associated with cancer development may either be specific for certain tumors or nonspecific, and involve multiple aberrations. Such chromosomal abnormalities are composed of amplification, deletion, or rearrangement of DNA in the form of a whole chromosome or its fragments. Tumors of the skull base are rare neoplasms that grow along various regions of the bottom part of the cranial vault and may be of intracranial or extracranial origin. A wide variety of benign and malignant tumors of endodermal, mesodermal, and epidermal origin may arise in this anatomical compartment, thus posing a considerable diagnostic challenge to the pathologist. These tumors may primarily arise in the skull base or invade the skull base locally from adjacent sites, such as the nasal cavity or paranasal sinuses. Significant progress has been made during the last decade in the understanding of the cytogenetic and molecular pathophysiology of various neoplasms, but cytogenetic information on most skull base tumors is scarce due to their rarity. Improved understanding of genetic mechanisms involved in tumorigenesis might help develop new diagnostic tools and define markers for estimating prognosis and for identifying patients at risk for recurrence. Genetic data may aid in the development of novel therapeutic modalities, including gene therapy and small molecules that directly affect abnormal genes or their products. The current chapter introduces the methods used for diagnosis and research of skull base neoplasms and reviews the current cytogenetic and molecular information that has emerged during the last decade on skull base tumors.
Fluorescent In Situ Hybridization and Spectral Karyotyping Since the introduction of conventional chromosomal banding during the 1970s, the technique has remained the working horse of cytogeneticists for screening of chromosomal aberrations found in human neoplasms. A new technical development based on fluorescence in situ hybridization (FISH) was added to our armamentarium of cancer research tools during the past decade. FISH evolved to solve the main problems of chromosomal banding: it was time consuming, error prone, and dependent upon highly qualified human resources. The main advantage of FISH is its ability to detect complex chromosomal deviations, which are often found in cancer specimens. Several methods utilizing multicolor– FISH (m-FISH) have been described, including multiplex– FISH (M-FISH), combined binary ratio–FISH (COBRA-FISH), rainbow cross-species color banding (Rx-FISH), and spectral karyotyping (SKY). These methods enabled visualization of all 24 human chromosomes, with unique chromosome-painting probes and a specific spectral color for each chromosome
GENETIC METHODS Conventional Chromosomal Analysis Conventional chromosomal staining is performed by Giemsa staining on primary short-term cultures. This analysis 63
64
Gil and Fliss
Figure 1
An ideogram of all 24 chromosomes in a normal human karyotype.
Figure 2 A representative karyotype from a conventional chordoma showing a normal 46,XY karyotype.
Chapter 3: Genetic Abnormalities of Skull Base Tumors
65
Figure 3 Chromosomal staining, using SKY, showing a normal female karyotype (46,XX).
pair. These probes interact with known distinctive sequences along the length of the chromosome. The approach combines Fourier spectroscopy, charge-coupled device imaging, and optical microscopy. Chromosome-specific composite libraries were generated by polymerase chain reaction (PCR) from flow-sorted human chromosomes conjugated to five different fluorescence dyes. Hybridization is carried out at 37◦ C in a humid chamber for 72 hours. The metaphases are captured and analyzed using a dedicated fluorescent microscopy equipped with a triple filter. In order to generate a classified spectral karyotype, the acquired spectral image is analyzed using SKYVIEWTM software. Figure 3 shows a typical chromosomal staining in SKY. This technique of chromosome labeling has a special value as an adjunct to classical karyotyping for confirmation of established translocations and in spotting additional translocations and insertions that were not identified by the banding technique. SKY can also aid in identifying the origin of a chromosomal fragment, such as double minutes. One powerful application of FISH is the use of locusspecific probes (LSPs) aimed at detecting particular DNA sequences present in a single chromosome level. These LSPs can serve as a tool to detect a known genomic deviation (i.e., deletions, amplifications, inversion, and translocation) that may sometimes be the only diagnostic evidence for a specific cancer. Commercially available molecular kits based on FISH are useful diagnostic tools in cases of neoplasms with low mitotic activity, sparse pathologic material, and nonfresh tissue samples.
Comparative Genomic Hybridization Both the conventional G-banding and SKY techniques pose several difficulties when dealing with tumor preparations. First, very few metaphase spreads can be obtained after cell culturing and their quality may be inadequate to allow recog-
nition of chromosomal banding. Second, these techniques require the utilization of fresh tissue samples whereas most tissue preparations are embedded and preserved as paraffin blocks. Third, the introduction of new chromosomal aberrations as an artifact of the culturing process that otherwise would not be a part of the tumor configuration cannot be ruled out. Therefore, another method free of these three drawbacks of conventional cytogenetics is required for our arsenal of cytogenetic research modalities. Comparative genomic hybridization (CGH) was developed to compensate for these problems encountered when looking for recurrent breakpoints in cancer cells DNA. CGH allows scanning of the entire genome of tumor cells in a single step, searching for gains or deletions of chromosomal material compared to normal tissue. The first step of CGH involves isolation of genomic DNAs from the tumor (test) and normal (reference) samples. Each DNA sample is subject to competitive hybridization to normal metaphase chromosomes. Labeling of the DNA is performed by using red and green fluorescent dyes for the tumor and normal samples, respectively. The final analysis is performed by comparing the differential ratio of red and green fluorescence along the longitudinal axis of each chromosome separately. The gains of chromosomal fragments in the tested sample compared to the reference are indicated by green and the deletions are indicated by red. Chromosomal regions that are equally distributed along the chromosome of the test as well as the reference tissue samples are indicated by yellow. Figure 4 shows a summary of DNA sequence copy number gains and losses in 16 specimens of malignant fibrous histiocytoma (MFH) analyzed by CGH. The main advantage of CGH is its ability to analyze numerous samples of tumors preserved in paraffin, allowing comparison studies with the clinical correlates of the tumor karyotypic features. The technique allows detection of recurrent aberrations of rare tumors such as those found in the skull base. Using the CGH technique, Sakabe et al. found
66
Gil and Fliss
Figure 4 Summary of DNA sequence copy number gains and losses in 16 specimens of malignant fibrous histiocytoma analyzed by CGH. Losses are shown on the left and gains are shown on the right. Each line represents a genetic aberration seen in one sample. Open boxes signify high-level amplifications of small chromosomal regions. Source: From Ref. 1.
a distinct amplification at 8p23.1 in MFH, leading to the detection of a new gene MASL1 whose expression is enhanced in MFH tumors bearing the 8p amplicon (1). The main limitation of CGH is its inability to detect balanced chromosomal rearrangements, which have a key role in the development of many types of cancer. Genomic imbalances that can be detected by CGH are found in approximately half of the human neoplasms (2).
Identification of Submicroscopic Genetic Deviations The cytogenetic methods described thus far involved chromosomal abnormalities at the microscopical level. Genetic abnormalities involved in carcinogenesis, however, may be submicroscopic in the presence of a normal karyotype. These changes cannot be demonstrated by cytogenetic methods and require molecular analysis. Submicroscopic genetic deviations can involve changes in the DNA, mRNA or protein level. Changes in DNA sequence are detected by PCR or Southern blotting, mRNA deviations by Northern blotting of PCR, and alterations in protein products by Western blotting. Molecular methods are currently used mainly for research, but they will soon provide novel approaches to assist physicians in detecting cancer, in tailoring the modality of treatment and in estimating a patient’s outcome during and after treatment.
Molecular Arrays The development of new methods for mass screening of gene expression, such as CGH-arrays, DNA-arrays, tissue microarrays, and protein-arrays, paved the way for the discovery of
complex cascades of molecular events leading to the development of cancer (3). These technologies allow scientists to monitor change in gene expression during tumor propagation and under different conditions. DNA-arrays allow largescale quantification of tens of thousands of genes in a single experiment. The platform for this technology is a solid stage (the target chip) on which an array of thousands of oligonucleotides (or cDNAs) is printed in a predetermined pattern. The construction of a DNA-array dedicated for human or cancer research can be commercially acquired or constructed according to the experiment prerequisite. Two main systems of DNA-arrays are commercially available to the scientist: the first is an array of cDNA clones robotically mottled on a solid platform, and the second is a technique that utilizes arrays of oligonucleotides directly synthesized on a solid chip. Each oligonucleotide contains 20 to 80 base pairs, which are a signature sequence of specific genes. Commercially available DNA chips contain 36,000 or 60,000 potential genes participating in murine or human cancer, respectively. The first stage is preparation of the tissue samples and array hybridization in the presence of an access target. In this step, fluorescent probes are derived from the specimens’ mRNAs by reverse transcription and hybridized in a complimentary fashion on the target array. The second step is image acquisition, in which the emitted fluorescence is detected by a camera and given a quantitative value for the expression of each gene (Fig. 5). The last and most complicated step of this method is data analysis during which tumors are segregated according to their gene expression profile. The amount of labeled cDNA that hybridizes with its target is proportional to its abundance in the original sample. Intensities are normalized and analyzed with powerful bioinformatic tools and correlated with the clinical characteristics of the tumor and the patient. A complimentary new technology for investigation of protein expression in cancer is antibody microarrays. The technique allows parallel investigation of thousands of proteins and screening of many markers known to participate in tumorigenesis. This method uses an array of different antibodies immobilized on a solid surface which interacts with fluorescence-labeled proteins. The tested protein can be directly labeled using a fluorophore (label-based) or indirectly labeled using a secondary antibody (sandwich) (4). Proteinarrays were used to detect protein expression in squamous cell carcinoma (SCC) of the head and neck as well as other malignancies utilizing laser capture microdissection (5,6). The versatile capability of protein-arrays allows simultaneous screening of thousands of candidate protein markers differentially expressed in distinct tumors and fishing those which are significantly correlated with the clinical status of the patient. The emergence of these powerful genetic tools will aid scientists to molecularly portrait the pathogenetic mechanism underlining the development of cancer. The clinical objectives of these modalities are to develop more accurate classification systems based on gene or protein expression profiles of tumors, to detect molecularly discrete subgroups of patients allowing prognosis stratification and tailoring specific therapy for a particular patient.
COMMON CHROMOSOMAL ABNORMALITIES IN CANCER As noted previously, common chromosomal aberrations found in benign or malignant neoplasms frequently involve change or rearrangement of genetic material in the form of chromosome number or chromosomal fragments.
Chapter 3: Genetic Abnormalities of Skull Base Tumors
67
Figure 5 Clustering of gene expression in three soft tissue tumors using GeneChip Arrays. (A) Synovial sarcoma, (B) liposarcoma, and (C) schwannoma. The samples were clustered using a set of 2600 known gene fragments. The tissue samples in the tree tumors are joined by very short branches when their gene-expression patterns are very similar; they are joined by increasingly longer branches as their similarity decreases. The color of each square represents the ratio of the gene expression in the indicated sample relative to the average signal of expression of all genes examined. Red indicates the gene expression above the median, green indicates expression below the median, and black indicates expression equal to the median. The intensity of the color reflects the magnitude of divergence from the median. Columns represent the indicated tissue sample; rows represent individual cDNAs. Source: From Ref. 102.
In the following section, we will describe some of the most important chromosomal abnormalities found in tumor specimens (Table 1). Chromosomal deletion involves the loss of a whole chromosome or its fragment (Fig. 6). A deletion that causes removal of a gene that suppresses cell growth and replication may trigger cancer. Such a gene that normally limits the growth of tumors is named tumor suppressor gene. Inactivation of a tumor suppression gene frequently involves mutation of one allele and deletion of a chromosomal region of the second allele. Chromosomal gain involves addition of a whole chromosome or its fragment to the normal 24 human chromosomes (Fig. 7). Deletions and gains of chromosomal fragments or loss of an entire chromosome are common in skull base carcinomas and are probably pathological consequential alterations. Such genomic loss can be detected using a restriction fragment length polymorphism analysis known as loss of heterozygosity (LOH). Translocations occur when two or more chromosomes exchange material (Fig. 8). Translocations induce rearrangement of genetic material that may lead to juxtaposition of genes, resulting in the formation of chimeric genes and abnormal production of protein affecting cell cycle
Table 1 Glossary of Cytogenetic Terms and Abbreviations p q t inv del i + −
dic r ins der
Arm of chromosome above centromere, generally a short arm of a chromosome. Arm of chromosome below centromere, generally a long arm of a chromosome. Translocation: occur when two or more chromosomes exchange material. Inversion: reversal of chromosomal material. Deletion: loss of chromosomal material. Isochromosome: symmetric chromosome composed of duplicated long or short arm with associated centromere. This symbol before a chromosome number indicates an additional whole chromosome and after a chromosome number and arm designation indicates an additional material on that arm. This symbol before a chromosome number indicates loss of a whole chromosome and after a chromosome number and arm designation indicates loss of material on that arm. Dicentric chromosome with two centromeres. Ring chromosome Insertion of extra material within a chromosome Derivative chromosomes: an abnormal chromosome resulting from structural rearrangement involving two or more chromosomes, generally of unbalanced nature.
68
Gil and Fliss
Figure 6 A karyotype from a poorly differentiated nonkeratinizing SCC of the skull base showing chromosomal deletion. The karyotype shows two deletions: del(6)(p21) and del(7)(p13) as a part of a complex karyotype (black arrows). Other aberrations are also seen.
and regulation. Genes whose activations contribute to the formation of cancer are known as oncogenes. A number of translocations discovered in human cancer led to the finding of new oncogenes and revealed the mechanism responsible for tumor development (7,8). Many specific translocations were found since the discovery of the first specific abnormality in human cancer t(9;22)(q34;q31) known as the Philadelphia chromosome (9). Translocations are involved in various lymphomas, leukemias, and sarcomas. Among skull base
A
B
Figure 7 A karyotype from a porocarcinoma of the temporal bone showing chromosomal gains. SKY analysis shows gains of chromosomes 6, 8, and 20.
Figure 8 A karyotype from a malignant peripheral nerve sheath tumor of the anterior skull base showing two balanced translocations. G-banding (A) and SKY analysis (B) show translocations involving (2;4)(q35;q31) and (12;X)(q24;q22).
Chapter 3: Genetic Abnormalities of Skull Base Tumors
tumors, alveolar rhabdomyosarcoma, Ewing sarcoma, and synovial sarcoma are characterized by specific translocations. Insertion of a chromosomal fragment to a foreign location or its inversion on the same chromosome may also change the location, resulting in the formation of chimeric genes and the formation of abnormal protein contributing to tumor growth. Gene amplification is defined as an increase in the number of copies of a gene which occurs by replication of the genomic DNA. The MYC and RAS families are genes that are most frequently found to be amplified in various skull base malignancies. Amplification of single or multiple genes can be identified by FISH. Double minutes are extrachromosomal circles of DNA containing up to 2 million base pairs, which form as a result of genetic instability and replicate autonomously. Gene amplification in the form of double-minute chromosomes may result in increasing oncogene products, contributing to malignant tumor formation. Figure 9 shows double-minute chromosomes present in dedifferentiated chordoma of the clivus. In this case, further chromosomal analysis using the SKY technique revealed the origin of DNA contained in these double-minute chromosomes.
69
GENETIC ABNORMALITIES OF SKULL BASE TUMORS Most tumors of epithelial origin [i.e., SCC, adenocarcinomas and sinonasal undifferentiated carcinomas (SNUC)] and several of the soft tissue sarcomas (i.e., osteosarcoma and chondrosarcoma) do not show specific chromosomal aberration. They often show complex karyotypes with numerous chromosomal deletions and gains. These changes may derive from a stepwise genetic transformations commonly seen in these tumors. In this scenario, a single genetic alteration may contribute to an increase in cell growth and replication, causing augmentation of the chromosomal aberrations contributing to tumor invasion and metastases, resulting in malignant transformation. Genetic transformations which can be found in a variety of benign tumors can induce cell proliferation without causing malignancy. The earliest work that systematically described the cytogenetic characteristics of 18 skull base tumors was reported by Gollin and Janecka, and about half of the tumors in their study failed to proliferate in culture (10). In a recent study, Gil and Fliss characterized the karyotype of 104 tumors that originated in the cranial base and successfully karyotyped 96% of them (11). The variation in results may be due to differences in the culture conditions or quality of the specimens. A significant increase in the abundance of chromosomal abnormalities is found in malignant skull base tumors compared to benign ones (50% and 20% of the specimens, respectively). Tables 2 and 3 summarize the specific and nonspecific chromosomal changes found in skull base neoplasms, respectively. In the following paragraphs, we present an overview of the cytogenetic properties found in skull base neoplasms.
Squamous Cell and Undifferentiated Carcinoma
C
Less than 4% of carcinomas arise in the skull base and paranasal sinuses (12). Cytogenetic features of these relatively rare tumors have yet to be fully characterized. Most sinonasal SCCs are well-to-moderately differentiated, with a positive stain to cytokeratin. Less frequently, poorly differentiated non-keratinizing SCC, adenocarcinoma, or SNUC, which lack the typical feature of conventional SCC, can arise from the sinonasal epithelium. Non-keratinizing SCC (i.e., transitional cell carcinoma, intermediate cell carcinoma, and schneiderian carcinoma) is considered a variant of SCC, whereas SNUC is a pathologically distinctive neoplasm which also lacks squamous differentiation. The latter tumor is composed of medium-sized undifferentiated cells, with occasional mixed neuroendocrine features. Three reports on five cytogenetically abnormal karyotypes of paranasal SCC have appeared in the English literature and one on undifferentiated carcinoma (10,13–15). Gil and Fliss have found that half of the non-keratinizing SCC specimens displayed an abnormal karyotype (11). Most of the chromosomal abnormalities involved several clones with complex karyotypes. SCC of the upper aerodigestive tracts frequently harbors a highly complex karyotype, as demonstrated in most paranasal SCC. Table 2 Diagnostic Chromosomal Changes in Skull Base Tumors Tumor
Figure 9 A karyotype from a dedifferentiated chordoma of the clivus showing complex karyotype with double minute chromosomes. (A) G-banding of metaphase spread showing polyploid complex karyotype with double minutes. (B) SKY analysis showing staining of chromosome 17. (C) G-banding of the double minute chromosomes.
Alveolar rhabdomyosarcoma Ewing sarcoma Synovial sarcoma Lipoma Myxoid liposarcoma
Chromosomal abnormality t(2;13)(q37;q14) or t(1;13)(p36;q14) t(11;22)(q24;q12) t(X;18)(p11.2;q11.2) t(12;?)(q14;?) t(12;16)(q13;p11)
70
Gil and Fliss
Table 3 Recurrent Non-Diagnostic Chromosomal Changes in Skull Base Tumors Tumor Malignant Adenocarcinoma Adenoid cystic carcinoma
Malignant peripheral nerve sheath tumor Esthesioneuroblastoma
Chromosomal abnormality +19p, +Xp22 t(5;9)(q15;q22), t(12;18)(p12;q12), t(1;2)(p36;q31), −9p22, del2(q32;q37) +17q , del(13)(q14−q21)
t(11;22)(q24;q12), t(21;22)(q22;q12), −1p, −3p/q, −8, −9p, −10p/q, +17p13, +20p, + 22q Malignant fibrous histiocytoma +1p, +4q12−21, +8p21, +8q24.1, +9q12−13, +12p11.2, +15q11.2−15, +19p, +7q32 Synovial sarcoma +5p15, 15p11, 19p13, −8, −9, −10p11, −18, −22q11 Osteosarcoma +1, −9, −10, −13, −17 Ewing sarcoma +8, +1q, der(16)t(1;16) Embryonal rhabdomyosarcoma +2, +7, +8, +12, +13, −11p15.5 Benign Meningioma −1p, −6q, −9p, −10, −14q, −18q, −22q12, +1q, +9q, +12q, +15q, +17q, +20q Pleomorphic adenoma t(3;8)(p21;q12) or t(9;12)(p13;q13) Chordoma −1p36, −3, −4, −10, −13, Lipoma del(13)(q12q22), der(6)(p21−23), der(11)(q13) Pituitary adenoma +12 Schwannoma/Neurofibroma −22q12.2
Chromosomal aberrations involving similar breakpoints to those found in the skull base were previously described in SCC of the oral cavity, larynx, hypopharynx, and nasopharynx (14,16,17). Deletion of chromosome Y is a common feature among skull base SCCs and is also frequently found in SCC of other parts of the upper aerodigestive tracts (16–18). Loss of the Y chromosome can be present in both normal and malignant cells and is known to increase with age in both situations. Therefore, the presence of a 45,X,-Y cell population in patients older than 75 years is most likely associated with advancing age, and should not be interpreted as a marker of the malignant clone. Recurrent chromosomal abnormalities involving 1p, 2q, 6q, 7p, 8q, and 12q are frequently found in paranasal SCC (Fig. 10). Although not specific, similar aberrations are found in SCC of the oral cavity, larynx, nasopharynx, and hypopharynx. Chromosomal abnormalities are also found in 60% of SNUC, half of them showing a complex karyotype (Fig. 11). In a recent work, Ariza et al. used CGH to detect recurrent chromosomal abnormalities in patients with sinonasal adenocarcinoma (19). Sinonasal adenocarcinoma differs from other carcinomas of the paranasal sinuses in its histological appearance. Adenocarcinoma has strong association with wood and leather manufacturing, and therefore serves as an excellent model for identification of carcinogenesis initiated by environmental mutagens. Common aberrations found in this tumor involve gain of chromosomal arms 19p and Xp22. Interestingly, gains of chromosomes 19p and Xp22 are also found in paranasal SNUC, suggesting a common mechanism for the development of sinonasal undifferentiated and adenocarcinoma.
Figure 10 Recurrent chromosomal aberrations in paranasal carcinomas. The y-axis shows the number of chromosomal aberrations found in the short (p) and long (q) chromosomal arms of 15 paranasal carcinomas. The x-axis indicates the chromosome number. The graph shows multiple chromosomal breakpoints that are characteristic of this tumor.
Salivary Gland Tumors of the Skull Base The most common salivary gland tumor originating in the skull base area is adenoid cystic carcinoma (ACC). Other tumors include acinic cell carcinoma, carcinoma expleomorphic adenoma, and polymorphous low-grade adenocarcinoma. Twenty-four cases of ACC cytogenetics were described previously in the English literature, and most of them involved the parotid and submandibular gland. Gil et al. analyzed the largest series of ACCs of skull base origin and found that one-third of them had an abnormal karyotype. Frequent chromosomal aberrations of skull base ACC include translocations of chromosome arms 5q15 and 9q22; 12p12 and 18q12; 1p36 and 2q31; as well as deletions of chromosome arms 9p22 and of 2(q32;q37). Other adnexal carcinoma invading the skull base may show a complex karyotype that
Figure 11 A karyotype of a sinonasal undifferentiated carcinoma of the anterior skull base showing a triploid complex karyotype: 60–69,XX,+der(X) add(X)(p22),+der(1) del(1)(q23),+der(1)del(1)(q32),+der(1) t(1;2)(q21; q13), +der(1) t (1; 12) (q21; q13), +2, +4, +5, +der(5)del(5)(q?15), +6, +der(6)i ? (p10),der(7)t(7; 15)(p12; q12), +der(8)t(8; 13)(q24; q11), +9, +12, −13, −13, +16, +16, +18,+19,+20,+der(20)add(20)(p13),+der (21)add(21)(p11), +22, +22, +1 ∼8mar[cp5]. Source: From Ref. 103.
Chapter 3: Genetic Abnormalities of Skull Base Tumors
71
Figure 12 A karyotype of a porocarcinoma of the lateral skull base showing a complex karyotype. (A) Sky analysis of the primary tumor. (B) Gbanding analysis of a positive neck lymph node metastasis from the same patient. Both specimens showed the same karyotype: 50,XX,inv(2)(p11;q13), +6,inv(7)(p13;p22),+8,+8,+20.
may also be evident in regional lymph node metastases to the neck (Fig. 12).
Clival Chordomas Chordoma is a rare malignant neoplasm derived from remnants of the embryonal notochord along the axial skeleton. It accounts for 1–4% of malignant osseous lesions, arising most frequently in the sacrococcygeal region and rarely in the spheno-occipital (clival) or vertebral regions. Most cytogenetic studies on chordomas have been on those of sacral or spinal origin. Sandberg and Bridge reported 29 cytogenetically abnormal cases of chordoma in a recent review of the literature (20). Fifteen of these cases involved the clivus. Previously reported frequent chromosomal aberrations found in chordomas involved loss of chromosomes 3, 4, 10, and 13, as well as structural rearrangement of chromosome arms 1p and 21q22 (20–22). Gil and Fliss found that one-half of the clival chordomas have an abnormal karyotype and show aberrations similar to sacral chordoma (23). Sawyer et al. reported a normal karyotype in all primary clival chordomas, whereas all recurrent tumors showed chromosomal aberrations (24). The authors therefore postulated that chromosomal aberrations occur late in the progression of chordoma and predominantly in recurrent cases. Genetic analysis using the LOH technique in patients with familial and sporadic chordomas has identified a potential region for a tumor suppression gene at 1p36 (25,26). A malignant variant of chordoma is dedifferentiated chordoma. This tumor shows elements of a conventional chordoma in association with areas of high-grade osteosarcoma. Cytogenetic analysis revealed a polyploid complex karyotype of 71 to 123 chromosomes with double minutes. Interestingly, polyploid karyotypes of 48 to 97 chromosomes with double minutes were also described in cases of highgrade osteosarcoma (26,27). In contrast, of all the previously described chordoma karyotypes, only one involved a polyploid karyotype with 72 chromosomes (28). Using the SKY technique, it was determined that the origin of the double minutes was in chromosome 17 (Fig. 9). Few genes involved in tumorigenesis are located on chromosome 17, among them are TP53, BRCA1, ERBB2, and NF1. Identification of the ex-
act area on chromosome 17 from which the double minutes originate awaits further studies. The growing complexity of the karyotype found in dedifferentiated chordoma is in keeping with the presently accepted view that a multistep process involving oncogenes, loss of tumor suppression genes, and other genetic alteration is cooperatively involved in anaplastic transformation (Fig. 13) (24,29). The pathogenetic evolution of the dedifferentiated chordoma and its relation with simple chordomas is also unknown. It may evolve from a common progenitor stem cell (i.e., the divergent theory) or from separate clones with distinct differentiation pathways (i.e., the collision tumor or convergent theory). If two clones are involved, one clone differentiates into a chordoma and the other (which is void of any differentiation potential) evolves into a high-grade sarcoma (30). Alternatively, it was suggested that the high-grade components of the tumor might emerge directly from the more indolent chordoma by a mechanism of dedifferentiation (31). The exact mechanism of dedifferentiation is not clear: it has been postulated that progression toward an anaplastic component could be regarded as a failure of differentiation rather than a reversal of differentiated cell to an embryonic, undifferentiated cell (32).
Neurogenic Skull Base Tumors Neurogenic tumors that originate in the skull base can be benign or malignant. Two of the most characterized benign tumors that originate in the skull base are schwannomas and neurofibromas. Neurofibroma type 2 (NF2) is an autosomal dominant disorder that is frequently associated with bilateral vestibular schwannomas along with other tumors in the central nervous system. The 100 kb NF2 tumor suppressor gene is located on chromosome arm 22q12.2, and gives rise to a 595 amino acid merlin protein (33,34). This protein is involved in either inter- or intracellular interactions which control cell motility, shape, and communication. Loss of function of the merlin protein results in tumor development due to loss of contact inhibition (35). Mutation of the NF2 gene can also lead to the development of other skull base tumors, such as meningiomas and ependymomas. Half of the patients suffering from NF2 disease have positive family history for the
72
Gil and Fliss
Primary clival chordoma
46,XY
Recurrent clival chordoma
Dedifferentiated clival chordoma
46,XX complex karyotype
Polyploid 71-123
Figure 13 A multistep model involving multiple genetic alterations suggested for the development of dedifferentiated chordoma. Three different cases of clival chordoma showing gradual change in karyotype: Primary chordoma showing normal karyotype (left), recurrent chordoma with a complex karyotype (middle), and dedifferentiated clival chordoma with polyploid karyotype (right). Arrowheads indicate chromosomal aberrations.
disorder. The rest are sporadic cases. More than 200 known mutations were identified in patients suffering from NF2 disease, where the more severe form is associated with a truncated protein. Among the common malignant skull base tumors of neuroectodermal origin are esthesioneuroblastomas, Ewing sarcoma, Askin tumor, and malignant peripheral nerve sheath tumor (MPNST). Conventional histopathological classification of these tumors may be complicated since these neoplasms contain similar morphological and cytological features. The common simple translocation found in Ewing sarcoma is t(11;22)(q24;q12), which is found in 78% of the affected patients. This chromosomal rearrangement can be found as a single translocation or as a part of a complex karyotype in which other chromosomes are involved. More than 90% of the tumors can be identified using 22q12 or 11q24 chromosome markers. Additional chromosomal changes found in Ewing sarcoma are trisomy 8 (44%), trisomy 1q (18%), and der(16)t(1;16). Gains or losses of chromosomes 2, 9, and 7 may be each found in 20% of these patients. Esthesioneuroblastoma is a relatively uncommon tumor that arises from the olfactory epithelium in the upper nasal cavity and often shows an intracranial extension. Esthesioneuroblastomas closely resemble Ewing sarcoma and other peripheral primitive neuroectodermal tumors and therefore present a diagnostic challenge to the pathologist. About 1000 cases have been reported in the world literature, but only 3 cases with a whole karyotype have been published (13,36). Two of these cases showed a complex hyperdiploid karyotypes with different numerical and structural aberrations. Several studies have suggested that the t(11;22)(q24;q12) or the t(21;22)(q22;q12) translocation, which is frequently present in peripheral primitive neuroectodermal tumors, including Ewing sarcoma, is also a characteristic finding of esthesioneuroblastomas. In contrast, others failed to find these translocations in esthesioneuroblastoma specimens (Fig. 14) (11,37–39). Some investigators have used CGH to screen for chromosomal breakpoints of esthesioneuroblastomas, and their findings suggested a characteristic pattern consisting of deletions of chromosomes 3p and overrepresentations of 17q (40,41). Other important alterations were deletions of 1p, 3p/q, 9p, 10p/q along with gains of 17p13, 20p, and 22q (41). Amplification of
chromosome 8 was also shown previously to exist in this tumor (11,38,42).Taken together, the recent cytogenetic works provide further evidence that esthesioneuroblastoma is not a part of the peripheral nerve sheath tumor (PNET) family. MPNST, also known as neurofibrosarcoma, malignant schwannoma, or neurogenic sarcoma, is a rare neoplasm of Schwann cell origin. Eighty-three cases of MPNSTs with chromosomal aberrations have been reported in the literature (43). Most of the MPNSTs had a complex karyotype with triploid or tetraploid clones (44). It was suggested that polyploid complex karyotype is a characteristic of MPNST, that it might be associated with high-grade tumors, and that it might serve as a discriminator between malignant and benign peripheral nerve tumors (which almost always display a near-diploid karyotype) (45). The discovery by Gil et al., of a simultaneous occurrence of a t(2;4)(q35;q31) and a t(X;12)(q22;q24) as the only chromosomal abnormalities in a high-grade MPNST, challenges this hypothesis (Fig. 8) (46). Other breakpoints frequently found in this tumor involve the long arm of chromosomes 2 and 4 (47). As stated before, the t(11;22) translocation is found in various neuroectodermal tumors (i.e., Ewing sarcoma esthesioneuroblastoma, peripheral neuroepithelioma, and Askin tumor), suggesting a common pathway for tumorigenesis in these neoplasms. Indeed, it was shown that various oncogenes are expressed in both esthesioneuroblastoma and Ewing sarcoma. Some of these oncogenes are c-myc, c-src, c-myb, and c-mil/raf (47). Analysis of the src protein levels in esthesioneuroblastoma, neuroepithelioma, and Ewing sarcoma demonstrated high protein levels and tyrosine kinase activity in these tumors relative to other neuroendocrine tumors (48). On the other hand, the c-myc oncogene is highly expressed in Ewing sarcoma but not in esthesioneuroblastoma (49). Further studies are warranted to show whether high expression of these oncoproteins is of etiologic significance. It is possible that multiple activated genes take part in cancerous transformation, either independently or in cooperation with one another.
Meningioma Meningioma was the first solid tumor to show characteristic cytogenetic abnormality, that is, loss of chromosome 22 (50).
Chapter 3: Genetic Abnormalities of Skull Base Tumors
73
Figure 14 A karyotype of a recurrent esthesioneuroblastoma of the anterior skull base showing a complex karyotype: 46,XX,−2,i(2)(q10),t(5;6)(q31; q26), −7,der(7)t(7; ?)(p22; ?), −8, −9,der(9)t(9; ?)(q34; ?), −11,der(12)(t(9:12)(q13;p13),−15,−17,+7mar. Red boxes showing the chromosomal aberrations.
Loss of heterozygosity on chromosome 22q is frequently observed in meningioma, leading to a bi-allelic inactivation of the NF2 tumor suppressor gene (51). After vestibular schwannomas, the second most common disease associated with mutation of the NF2 gene is meningioma (52). In the pediatric population, approximately 40% of the meningiomas are associated with NF2, which must be ruled out in any child with meningioma especially if it is multifocal. Other less common genetic syndromes associated with meningioma are non-NF2-associated meningiomas, Cowden syndrome, Gorlin nevoid basal cell syndrome, Li-Fraumeni syndrome, Turcot/Gardener syndrome, and von Hippel-Lindau disease. Previous studies have found numerous cytogenetic alterations associated with the progression of meningioma as well as with anaplastic and atypical variant. These include dicentric and ring chromosomes; loss of chromosome arms 1p, 6q, 9p, 10, 14q, and 18q; as well as gains of 1q, 9q, 12q, 15q, 17q, and 20q (53,54). A model of meningioma progression based on cytogenetic analysis is shown in Figure 15. As shown in Table 4, a world health organization (WHO) grade I menin-
gioma shows loss of chromosome arm 22q in 58% of the cases, a grade II tumor shows notable loss of 1p (76%) and 22q (71%), whereas a grade III tumor (anaplastic meningioma) is associated with −6q (53%), −10q (68%) and −14q (63%). Olfactory groove meningioma (OGM) is a slowgrowing meningioma that originates in the cribriform area and frequently achieves a large size before detection. In some cases, the tumor may involve surrounding structures, such as optic nerves and cerebral arteries, complicating its removal. Gil and Fliss analyzed a specific subgroup of meningiomas originating in the olfactory groove and found that 25% of them showed an abnormal karyotype, including breakpoint of chromosome arm 22q11 (Fig. 16). Similar abnormalities of OGMs were described by others (10,55). The presence of
Loss on: 22q, 18p
Table 4
Frequency of Cytogenetic Alterations in Meningiomas
WHO classification I II III
Normal
Breakpoint
(%)
22q 1p 22q − 6q −10q −14q
58 76 71 53 68 63
Gain on:1q, 9q, 12q, 15q, 17q, 20q Loss on: 1p, 6q, 10q, 14q, 18q, 22q
Benign meningioma
WHO I
Gain on 17q Loss on 6q: 9p, 10q: 14q
Atypical meningioma
WHO II
Anaplastic meningioma
WHO III
Figure 15 A model of meningioma progression based on cytogenetic analysis.
74
Gil and Fliss
Figure 16 Abnormal karyotypes of OGM. A representative hypodiploid karyotype of OGM showing dic(19:22)(p13:q11). The arrow points to the translocation between chromosomes 19p13 and 22q11 that generated a dicentric chromosome.
dic(19;22)(p13;q11) as well as the loss of chromosome arm 22q is consistent with aberrations noted in meningiomas of other areas (56–58). These findings suggest that similar genetic mechanisms take part in the development of meningiomas located in the olfactory groove as well as in other intracranial regions.
Juvenile Nasopharyngeal Angiofibroma Juvenile nasopharyngeal angiofibroma (JNA) is a benign tumor that occurs exclusively in adolescent and young adult males. This is a locally invasive fibrovascular tumor that originates in the sphenopalatine foramen area. The three works that used CGH analysis of JNA were all from the same laboratory (59–61). DNA gains were observed on chromosomes 3q, 4q, 5q, 6q, 7q, 8q, 12p, 12q, 13q, 14q, 18q, 21q, and X, and DNA losses were on chromosomes 17, 19p, 22q, and Y. The deletions of chromosome 17 included regions for the tumor suppressor gene P53 as well as the Her-2/neu oncogene. In 71% of their cases, gene losses were detected for both genes. However, comparison of p53 and Her-2/neu mRNA levels in laser microdissected endothelial and stromal cells were not conclusive to answer the question of the tumor cell of origin in JNA. A significant loss of the chromosome Y and gain of chromosome X was observed in 86% and 71% of the tumors, respectively. A gain of chromosome X leads to androgen receptor gene gain may suggest that JNA is an androgen-dependent tumor. This is supported by the finding that beta-catenin known to be over-expressed in JNAs acts as a co-activator of the androgen receptor. Gil et al. performed the first full karyotype characterization of JNA using the G-banding cytogenetics method (11). All 12 specimens that had been confirmed histologically as JNA showed a normal male karyotype. This result could represent the true chromosomal pattern of this tumor, similar to other benign tumors, particularly those of soft tissue and bone. There is also the possibility that a genetic change was present in some of the tumors, but that this change was submicroscopic and not detectable by a classical cytogenetic method. Another explanation for a normal karyotype is that the findings of G-banding or SKY analysis may represent a normal stromal cell proliferation in culture rather than tumor cell growth, and therefore represent the constitutional karyotype of the patient.
Pleomorphic Adenoma Pleomorphic adenoma is the most common tumor of the salivary glands. Rarely, the tumor may invade the parapharyngeal space, infratemporal fossa, and intracranial compartment. Such tumors are locally aggressive and their biological
behavior may be different from that of ordinary pleomorphic adenomas of the parotid gland. We applied cytogenetic methods in order to delineate the spectrum of chromosomal rearrangements of giant pleomorphic adenomas, which involved the lateral skull base area and found that 80% of them showed an abnormal diploid karyotype. Two major chromosomal subgroups had been recognized earlier in pleomorphic adenoma: one of them was characterized by structural rearrangements of 8q12 (62). This chromosomal rearrangement may be present in up to 60% of the cases with an abnormal karyotype. Translocations involving 8q12 may result in replacement of the promoter region of the PLAG1 gene by that of another gene (63,64). One such candidate gene located on 6q12 is PRL-1 and its expression is elevated in a number of tumor cell lines (65). PRL-1 encodes a 20-kDa protein with an eight-amino-acid consensus protein tyrosine phosphatase active site. The gene is important in normal cellular growth and can promote invasion activity and metastasis (66). Some of the skull base pleomorphic adenomas show a complex karyotype with ring chromosomes and double minutes (67). Such chromosomal aberrations are infrequent in this tumor, and their presence may have contributed to the aggressive behavior of the ones which invade the skull base.
Paraganglioma Familial paraganglioma is inherited in an autosomal dominant manner, with maternal imprinting. In this disease, an individual harboring an abnormal gene can bear a child affected with the disease only if the transmitting parent is the father (69). Genetic analysis of patients with hereditary paraganglioma revealed three genes whose mutations result in the development of tumors. Analysis of two families with hereditary paraganglioma located the underline gene at 11q23 (69,70). Physical analysis of this hereditary paraganglioma type 1 (PLG1) locus revealed a candidate gene encoding the enzyme succinate dehydrogenase D (SDHD). The SDHD gene encodes the 159 amino acid subunit of cytochrome b, located in the mitochondrial complex II. Various mutations of this polypeptide were found in patients with hereditary paraganglioma, responsible for 50% of the familial cases (71,72). Further studies of the SDHD protein complex revealed other genes encoding for this subunit of the mitochondrial complex, including SDHC (PLAG3) located at 1q21 and SDHB (PLAG4) located at 1p36 (73,74). Mutations of the SDHD and SDHB genes are also found in 8–36% of the patients with sporadic paragangliomas (75). The mechanism by which dysfunction of the mitochondrial complex induces tumors is unknown; however, it is conceivable that hypoxic stimulation due to chronic hypoxemia may give rise to proliferation of paraganglioma cells. Astrom et al. have recently demonstrated that higher altitudes and nonsense/splicing mutations are associated with phenotypic severity in patients with hereditary paraganglioma type 1 (76). Their result supports the hypothesis that SDHD mutations impair oxygen sensing. Knowledge of the genetic mechanisms involved in evolution of paragangliomas enabled the development of screening tests for potential candidates in families with hereditary paraganglioma (76).
Sarcomas Soft tissue sarcomas are infrequently found in the head and neck region in the mature population. These tumors are more common in childhood where up to 40% of them occur in the head and neck. Cytogenetic characteristics of soft tissue sarcomas are well established, but limited data exist on soft tissue sarcoma of the cranial base and head
Chapter 3: Genetic Abnormalities of Skull Base Tumors
75
and neck because of the rarity of this tumor. Histologically, it is difficult to distinguish between skull base soft tissue tumors based on conventional pathological and cytological measures. For example, distinguishing between Ewing sarcoma, embryonal rhabdomyosarcoma, small cell osteogenic sarcoma, and lymphoma may be challenging (78). Rhabdomyosarcoma (RMS) is the most common soft tissue sarcoma of childhood, and also the most common skull base malignant tumor in the pediatric population (79). RMS is divided into two major histologic subtypes, embryonal (ERMS) and alveolar (ARMS). ERMS typically occurs in children <10 years of age and most frequently in the head and neck or skull base regions. In contrast, ARMS most commonly occurs in adolescents and young adults and typically in the trunk and extremities. The prognosis of ERMS is generally favorable compared to ARMS which portends a poor prognosis. Chromosomal analyses of ARMS revealed a specific translocation associated only with this disease. The t(2;13)(q35;q14) translocation is found in approximately 70% of patients with ARMS (Fig. 17). Other studies have also shown a t(1;13)(p36;q14) variant translocation which may be identified in the minority of patients. Physical mapping and cloning studies demonstrated that the loci on chromosomes 2 and 1 rearranged by the t(2;13) and t(1;13) are PAX3 and PAX7, respectively (80,81). The chromosome 13 locus juxtaposed with both PAX3 and PAX7 is FKHR (FOX01A), which encodes a novel member of the fork head transcription factor family (82). The translocations break within intron 7 of PAX3 or PAX7 and intron 1 of FKHR and thus create two chimeric genes on the derivative chromosomes. A series of molecular studies demonstrated that these gene fusion events result in alterations at the level of protein function, gene expression, and subcellular localization (82). In contrast to the specific translocations found in ARMS, approximately 75% of ERMS cases have a complex karyotype. The most frequent chromosomal abnormalities of ERMS are translocation breakpoints in the 1p11-q11 region; gains of chromosomes 2, 7, 8, 12, and 13; and losses at chromosome 11p15.5 (83, 84). Gil et al. found that ERMS of the skull base are associated with translocations involving 13q14, which is a frequent finding in this tumor. Other aberrations of skull base ERMS include gains of chromosomes 5, 8, 9, and 12 (11). Synovial sarcoma is a highly malignant tumor infrequently involving the base of skull. It is difficult to distinguish from fibrosarcoma, solitary fibrous tumor, leiomyosarcoma, MPNST, and clear cell sarcoma or from soft tissue tu-
mors, such as esthesioneuroblastoma, rhabdomyosarcoma, and Ewing sarcoma (when it exhibits poorly differentiated features). The characteristic cytogenetic features of synovial sarcoma were fist described by Sandberg in 1986, who described the pathognomonic translocation t(X;18)(p11.2;q11.2) (85,86). Since then, serial cytogenetic examinations of synovial sarcomas established the presence of t(X;18) in more than 90% of the cases, and as the sole chromosomal abnormality in one-third of them (87). Until the publication of this book, nearly all studies have indicated that the (X;18) translocation is found almost exclusively in monophasic and biphasic synovial sarcoma. There is still heated debate on whether other soft tissue tumors, including MPNSTs, show similar translocation (88). In essence, the general consensus is that t(X;18) is characteristic of synovial sarcoma and no other tumor. The (X;18) translocation results in the fusion of two genes, the SYT gene at 18q11.1 and the SSX1 or SSX2 gene at Xp11.2 (89,90). The syt protein is a transcriptional activator located in the cell nucleus, whereas the ssx proteins are thought to act as transcriptional suppressors (91). Survival analysis of the two SYT–SSX fusion types showed that the SSX1 type is associated with shorter metastasis-free survival and poor prognosis compared with the SSX2 variant (92). A recent work by Gil et al. performed on over 100 cases of skull base neoplasms, established the diagnosis of synovial sarcoma in two patients, due to the presence of t(18;X)(q11;p11) (Fig. 18) (11). Other abnormalities, which were previously described in this tumor, including gains of 5p15, 15p11, and 19p13 and losses of 8, 9, 10p11, 18, and 22q11, are also present in synovial sarcoma of the skull base. These data confirmed that t(18;X) might be used for the diagnosis of synovial sarcoma originating in the skull base. MFH is the most common malignant soft tissue tumor in adults. This is also one of the most common sarcomas arising in the skull base. No specific chromosomal anomaly has been reported in MFH and most tumors analyzed thus far showed multiple numerical and structural abnormalities. CGH studies of MFH have found frequent gains at 4q1221, 8p21, 8q24.1, 9q12-13, 12p11.2, and 15q11.2-15 (1). Recurrent breakpoints were also found in 1p36, 1q11, 1q21, 3p12, 11p11, 17p11, and 19p13 (93). It was suggested that a +19p marker as well as gains of 1p and 7q32 might serve as an indicator for aggressive clinical course and dismal prognosis (94,95).
Figure 17 SKY analysis of alveolar rhabdomyosarcoma showing the t(2;13)(q35;q14) translocation. This rearrangement is found in approximately 70% of patients with ARMS.
Figure 18 Translocation involving (X;18) is found in more than 90% of the cases of synovial sarcoma, in third of them as the sole chromosomal abnormality.
76
Gil and Fliss
Osteosarcoma is an uncommon tumor of skeletal origin that may involve the skull base and craniofacial area. In children, this is the most common nonhematologic malignant tumor of bone, comprising more than half of the cases of bone malignancy among the pediatric population. Most osteosarcomas are high-grade tumors with propensity to metastasize, resulting in a less than 10% long-term disease-free survival. Unlike synovial sarcoma or ARMS, osteosarcoma is not associated with specific chromosomal abnormalities. Intensive genetic research has shown that genetic alterations in osteosarcoma include multiple, sequential changes involving both over-expression of oncogenes and inactivation of tumor suppression genes (96). A common mechanism believed to participate in the oncogenesis of osteosarcoma is the inactivation of the retinoblastoma gene (RB1) and TP53, leading to impaired regulation of cell proliferation (97). Other oncogenes that were shown to be over-expressed in osteosarcomas are FOS, MYC, MET, SAS, GLI, and ERBB2. The most common cytogenetic feature in osteosarcoma is a complex karyotype which resembles that of chondrosarcoma. Chromosomal abnormalities are found in 70% of the cases, ranging from haploid to hexaploid karyotype. The most frequent structural rearrangements found in osteosarcoma involve gain of chromosome 1 and loss of chromosomes 9, 10, 13, and 17. Other structural abnormalities, such as ring chromosomes, homogeneously staining regions, or double minutes, are frequently seen in conventional osteosarcoma. In a recent study, Gil and Fliss found microscopic genetic changes in 66% of skull base osteosarcomas (11). The majority of these cases were characterized by complex chromosomal abnormalities, multiple clones, and cell-to-cell variations. Cytogenetic abnormalities included complex karyotypes with frequent rearrangements of chromosomes 1, 2, 3, and 17. Single karyotypic changes were not found in skull base osteosarcoma cases.
FUTURE DIRECTIONS IN SKULL BASE GENETICS Gene Therapy Our armamentarium for the treatment of most aggressive skull base tumors lacks an effective modality for significantly improving long-term patient survival. The last decade has witnessed the development of gene therapy for a range of diseases, including inherited metabolic disorders and cardiovascular diseases. This approach has been explored extensively in the field of cancer therapy, resulting in the recent licensing of gene therapy for the routine treatment of head and neck cancer in China. Patients with skull base neoplasms may especially benefit from gene therapy because many of these tumors have already infiltrated cranial nerves and brain parenchyma at the time of diagnosis. The first goal of gene therapy is to replace a defective tumor suppression gene with a normal gene, reestablishing normal cell growth at the molecular level. Alternatively, shutting down an oncogene responsible for tumorigenesis will decrease its level of expression and will lead to the reduction of tumor growth. Unfortunately, many of the skull base malignancies lack a single candidate gene involved in tumorigenesis. Nevertheless, recent studies of benign tumors with a single gene mutation, including meningioma and schwannoma, showed promising results of gene therapy by decreasing tumor growth in vitro (98). Viruses are currently considered the most effective mean for gene delivery. A variety of virus vectors (i.e., adenovirus, vaccinia virus, and retroviruses) have been employed to deliver genes to cells, providing a transient or permanent
transgene expression. Modification of the virus envelope will enable the effective and safe use of systemic delivery methods of these viruses. Viral oncolytic therapy has also been studied extensively as a novel anticancer strategy. Whereas normal cells resist replication, tumor cells have an impaired antiviral response that sensitizes them to oncolytic viruses. Alone or in combination with chemotherapy or radiotherapy, viral oncolytic therapy uses the natural cytotoxicity of viruses to kill tumor cells exclusively. Among the most explored oncolytic viruses are the herpes simplex virus, adenovirus, and vaccinia virus. A number of these oncolytic viruses have the potential to become powerful anticancer agents, especially against advanced solid tumors. It has been 10 years since the beginning of the first clinical trial in which an oncolytic virus was administered to cancer patients. Since then, oncolytic viruses from five different species have been taken to phases I and II clinical trials in over 300 cancer patients. Importantly, the successful development of systemic administration of oncolytic viruses will extend the range of tumors that can be treated using this novel treatment modality and hopefully will allow effective and safe treatment of patients with advanced skull base tumors.
Skull Base Cytogenetics from Bench to Bedside and Back The ability to detect chromosomal aberrations and subtle molecular genetic abnormalities is important when considering cancer therapy and prevention. The emergent cytogenetic data established the development of diagnostic tests utilizing molecular approaches and allowed the development of various grading systems for assessing the diagnosis and prognosis of patients with cancer. The identification of specific translocations can be accomplished by using conventional cytogenetic methods on fresh tissues. New commercially available kits based on reverse transcriptase–polymerase chain reaction (RT-PCR) are a relatively new method for identifying specific cytogenetic abnormalities (99,100). Utilizing RTPCR–based diagnostic tests allows rapid, reliable, and inexpensive examinations of biopsies and fine-needle aspirates of skull base neoplasms in fresh or paraffin-embedded tissue. For example, the development of RT-PCR–based assay for synovial sarcoma showed 94–100% accuracy in a study of paraffin-embedded specimens (101).
SUMMARY Previous studies have demonstrated the prognostic value of cytogenetic data in head and neck epithelial and soft tissue tumors. This chapter reviewed the value of cytogenetic information for the diagnosis of soft tissue tumors, such as synovial sarcoma, ARMS and Ewing sarcoma. Some of the data presented here may serve as a starting point for further investigation on the biological behavior and tumorigenesis of several other skull base tumors, including paranasal carcinomas. To date, the extent of the cytogenetic data on most skull base neoplasms is too limited to furnish a precise association between karyotype and prognosis. The significance of many of the breakpoints discussed here awaits further clarification. Identification of other novel molecular alterations is necessary to improve the diagnosis of these complex tumors and to provide reliable molecular systems for predicting the prognosis and response to treatment of affected patients. Whether genetic data will become an essential adjunct to surgery and pathology in the treatment of skull base tumors awaits further studies.
Chapter 3: Genetic Abnormalities of Skull Base Tumors REFERENCES 1. Sakabe T, Shinomiya T, Mori T, et al. Identification of a novel gene, MASL1, within an amplicon at 8p23.1 detected in malignant fibrous histiocytomas by comparative genomic hybridization. Cancer Res. 1999;59:511–515. 2. Kallioniemi OP, Kallioniemi A, Sudar D, et al. Comparative genomic hybridization: A rapid new method for detecting and mapping DNA amplification in tumors. Semin Cancer Biol. 1993;4:41–46. 3. Bertucci F, Houlgatte R, Nguyen C, et al. Gene expression profiling of cancer by use of DNA arrays: How far from the clinic? Lancet Oncol. 2001;2:674–682. 4. Haab BB. Antibody arrays in cancer research. Mol Cell Proteomics. 2005;4:377–383. 5. Madoz-Gurpide J, Wang H, Misek DE, et al. Protein based microarrays: A tool for probing the proteome of cancer cells and tissues. Proteomics. 2001;1:1279–287. 6. Knezevic V, Leethanakul C, Bichsel VE, et al. Proteomic profiling of the cancer microenvironment by antibody arrays. Proteomics. 2001;1:1271–1278. 7. Solomon E, Borrow J, Goddard AD. Chromosome aberrations and cancer. Science. 1991;254:1153–1160. 8. Rabbitts TH. Chromosomal translocations in human cancer. Nature. 1994;372:143–149. 9. Tkachuk DC, Westbrook CA, Andreeff M, et al. Detection of bcr-abl fusion in chronic myelogeneous leukemia by in situ hybridization. Science. 1990;250:559–562. 10. Gollin SM, Janecka IP. Cytogenetics of cranial base tumors. J Neurooncol. 1994;20:241–254. 11. Gil Z, Orr-Urtreger A, Goldstein M. et al. Cytogenetic analysis of 101 skull base tumors. Head Neck, 2008;30:567–581. 12. Ganly I, Patel SG, Singh B, et al. Craniofacial resection for malignant paranasal sinus tumors: Report of an International Collaborative Study. Head Neck. 2005;27:575– 584. 13. Jin Y, Mertens F, Arheden K, et al. Karyotypic features of malignant tumors of the nasal cavity and paranasal sinuses. Int J Cancer. 1995;60:637–641. 14. Jin YS, Higashi K, Mandahl N, et al. Frequent rearrangement of chromosomal bands 1p22 and 11q13 in squamous cell carcinomas of the head and neck. Genes Chromosomes Cancer. 1990;2:198–204. 15. Zaslav AL, Stamberg J, Steinberg BM, et al. Cytogenetic analysis of head and neck carcinomas. Cancer Genet Cytogenet. 1991;56:181–187. 16. Jin Y, Mertens F, Mandahl N, et al. Chromosome abnormalities in eighty-three head and neck squamous cell carcinomas: Influence of culture conditions on karyotypic pattern. Cancer Res. 1993;53:2140–2146. 17. Jin C, Jin Y, Wennerberg J, et al. Nonrandom pattern of cytogenetic abnormalities in squamous cell carcinoma of the larynx. Genes Chromosomes Cancer. 2000;28:66–76. 18. Jin Y, Mertens F, Jin C, et al. Nonrandom chromosome abnormalities in short-term cultured primary squamous cell carcinomas of the head and neck. Cancer Res. 1995;55:3204– 3210. 19. Ariza M, Llorente JL, Alvarez-Marcas C, et al. Comparative genomic hybridization in primary sinonasal adenocarcinomas. Cancer. 2004;100:335–341. 20. Sandberg AA, Bridge JA. Updates on the cytogenetics and molecular genetics of bone and soft tissue tumors: Chondrosarcoma and other cartilaginous neoplasms. Cancer Genet Cytogenet. 2003;143:1–31. 21. Scheil S, Bruderlein S, Liehr T, et al. Genome-wide analysis of sixteen chordomas by comparative genomic hybridization and cytogenetics of the first human chordoma cell line, U-CH1. Genes Chromosomes Cancer. 2001;32:203– 211. 22. Tallini G, Dorfman H, Brys P, et al. Correlation between clinicopathological features and karyotype in 100 cartilaginous and chordoid tumors. A report from the Chromosomes and
23.
24.
25.
26.
27. 28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
77
Morphology (CHAMP) Collaborative Study Group. J Pathol. 2002;196:194–203. Gil Z, Fliss DM, Voskoboinik N, et al. Cytogenetic analysis of three variants of clival chordoma. Cancer Genet Cytogenet. 2004;154:124–130. Sawyer JR, Husain M, Al-Mefty O. Identification of isochromosome 1q as a recurring chromosome aberration in skull base chordomas: A new marker for aggressive tumors? Neurosurg Focus. 2001;10:1–6. Miozzo M, Dalpra L, Riva P, et al. A tumor suppressor locus in familial and sporadic chordoma maps to 1p36. Int J Cancer. 2000;87:68–72. Dressler LG, Duncan MH, Varsa EE, et al. DNA content measurement can be obtained using archival material for DNA flow cytometry. A comparison with cytogenetic analysis in 56 pediatric solid tumors. Cancer. 1993;72:2033–2041. Ozisik YY, Meloni AM, Peier A, et al. Cytogenetic findings in 19 malignant bone tumors. Cancer. 1994;74:2268–2275. Gibas Z, Miettinen M, Sandberg AA. Chromosomal abnormalities in two chordomas. Cancer Genet Cytogenet. 1992;58:169–173. Mitelman F, Johansson B, Mandahl N, et al. Clinical significance of cytogenetic findings in solid tumors. Cancer Genet Cytogenet. 1997;95:1–8. Tetu B, Ordonez NG, Ayala AG, et al. Chondrosarcoma with additional mesenchymal component (dedifferentiated chondrosarcoma). II. An immunohistochemical and electron microscopic study. Cancer. 1986;58:287–298. Ptaszynski K, Ramesh KH, Cannizzaro LA. Clonal chromosome aberrations with monosomy of chromosome 8 in a case of grade III chondrosarcoma. Cancer Genet Cytogenet. 1997;97:60–63. Casorzo L, Chiecchio L, Pisacane A, et al. Cytogenetic findings in a case of dedifferentiated chondrosarcoma. Cancer Genet Cytogenet. 2003;144:61–64. Trofatter JA, MacCollin MM, Rutter JL, et al. A novel moesin-, ezrin-, radixin-like gene is a candidate for the neurofibromatosis 2 tumor suppressor. Cell. 1993;75:826. Rouleau GA, Merel P, Lutchman M, et al. Alteration in a new gene encoding a putative membrane-organizing protein causes neuro-fibromatosis type 2. Nature. 1993;363:515–521. Gutmann DH, Sherman L, Seftor L, et al. Increased expression of the NF2 tumor suppressor gene product, merlin, impairs cell motility, adhesionand spreading. Hum Mol Genet. 1999;8:267–275 Whang-Peng J, Freter CE, Knutsen T, et al. Translocation t(11;22) in esthesioneuroblastoma. Cancer Genet Cytogenet. 1987;29:155–157. Kumar S, Perlman E, Pack S, et al. Absence of EWS/FLI1 fusion in olfactory neuroblastomas indicates these tumors do not belong to the Ewing’s sarcoma family. Hum Pathol. 1999;30:1356–1360. Mezzelani A, Tornielli S, Minoletti F, et al. Esthesioneuroblastoma is not a member of the primitive peripheral neuroectodermal tumour-Ewing’s group. Br J Cancer. 1999;81:586–591. Castaneda VL, Cheah MS, Saldivar VA, et al. Cytogenetic and molecular evaluation of clinically aggressive esthesioneuroblastoma. Am J Pediatr Hematol Oncol. 1991;13:62–70. Bockmuhl U, You X, Pacyna-Gengelbach M, et al. CGH pattern of esthesioneuroblastoma and their metastases. Brain Pathol. 2004;14:158–163. Riazimand SH, Brieger J, Jacob R, et al. Analysis of cytogenetic aberrations in esthesioneuroblastomas by comparative genomic hybridization. Cancer Genet Cytogenet. 2002;136:53–57. Van Devanter DR, George D, McNutt MA, et al. Trisomy 8 in primary esthesioneuroblastoma. Cancer Genet Cytogenet. 1991;57:133–136. Schrock E, du MS, Veldman T, et al. Multicolor spectral karyotyping of human chromosomes. Science. 1996;273:494– 497. Mitelman F, Johansson B, Mertens F eds. Mitelman database of chromosome aberrations in cancer. Available at: http:// cgap.nci.-nih.gov/Chromosomes/Mitelman. Accessed 2002.
78
Gil and Fliss
45. Mertens F, Dal Cin P, De WI, et al. Cytogenetic characterization of peripheral nerve sheath tumours: A report of the CHAMP study group. J Pathol. 2000;190:31–38. 46. Gil Z, Fliss DM, Voskoboimik N, et al. Two novel translocations, t(2;4)(q35;q31) and t(X;12)(q22;q24), as the only karyotypic abnormalities in a malignant peripheral nerve sheath tumor of the skull base. Cancer Genet Cytogenet. 2003;145:139–143. 47. Stephenson CF, Bridge JA, Sandberg AA. Cytogenetic and pathologic aspects of Ewing’s sarcoma and neuroectodermal tumors. Hum Pathol. 1992;23:1270–1277. 48. Bolen JB, Rosen N, Israel MA. Increased pp60c-src tyrosyl kinase activity in human neuroblastomas is associated with amino-terminal tyrosine phosphorylation of the src gene product. Proc Natl Acad Sci USA. 1985;82:7275– 7279. 49. Thiele CJ, McKeon C, Triche TJ, et al. Differential protooncogene expression characterizes histopathologically indistinguishable tumors of the peripheral nervous system. J Clin Invest. 1987;80:804–811. 50. Zang KD. Meningioma: A cytogenetic model of a complex benign human tumor, including data on 394 karyotyped cases. Cytogenet Cell Genet. 2001;93:207–220. 51. Ruttledge MH, Sarrazin J, Rangaratnam S, et al. Evidence for the complete inactivation of the NF2 gene in the majority of sporadic meningiomas. Nat Genet. 1994;6:180–184. 52. Perry A, Gutmann DH, Reifenberger G. Molecular pathogenesis of meningiomas. J Neurooncol. 2004;70:183–202. 53. Lamszus K, Kluwe L, Matschke J, et al. Allelic losses at 1p, 9q, 10q, 14q, and 22q in the progression of aggressive meningiomas and undifferentiated meningeal sarcomas. Cancer Genet Cytogenet. 1999;110:103–110. 54. Weber RG, Bostrom J, Wolter M, et al. Analysis of genomic alterations in benign, atypical, and anaplastic meningiomas: Toward a genetic model of meningioma progression. Proc Natl Acad Sci USA. 1997;94:14719–14724. 55. Ketter R, Henn W, Feiden W, et al. Nasoethmoidal meningioma with cytogenetic features of tumor aggressiveness in a 16-year-old child. Pediatr Neurosurg. 2003;39:190–194. 56. Rey JA, Bello MJ, de Campos JM, et al. Chromosomal involvement secondary to -22 in human meningiomas. Cancer Genet Cytogenet. 1988 15;33:275–290. 57. Sawyer JR, Husain M, Pravdenkova S, et al. A role for telomeric and centromeric instability in the progression of chromosome aberrations in meningioma patients. Cancer. 2000;88:440–453. 58. Sawyer JR, Husain M, Lukacs JL, et al. Telomeric fusion as a mechanism for the loss of 1p in meningioma. Cancer Genet Cytogenet. 2003;145:38–48. 59. Schick B, Veldung B, Wemmert S, et al. p53 and Her-2/neu in juvenile angiofibromas. Oncol Rep. 2005;13:453–457. 60. Schick B, Brunner C, Praetorius M, et al. First evidence of genetic imbalances in angiofibromas. Laryngoscope. 2002;112:397–401. 61. Schick B, Rippel C, Brunner C, et al. Numerical sex chromosome aberrations in juvenile angiofibromas: Genetic evidence for an androgen-dependent tumor? Oncol Rep. 2003;10:1251–1255. 62. Bullerdiek J, Wobst G, Meyer-Bolte K, et al. Cytogenetic subtyping of 220 salivary gland pleomorphic adenomas: Correlation to occurrence, histological subtype, and in vitro cellular behavior. Cancer Genet Cytogenet. 1993;65:27–31. 63. Kas K, Voz ML, Roijer E, et al. Promoter swapping between the genes for a novel zinc finger protein and beta-catenin in pleiomorphic adenomas with (3;8)(p21;q12) translocations. Nat Genet. 1997;15:170–174. 64. Voz ML, Astrom AK, Kas K, et al. The recurrent translocation t(5;8)(p13;q12) in pleomorphic adenomas results in upregulation of PLAG1 gene expression under control of the LIFR promoter. Oncogene. 1998;16:1409–1416. 65. Diamond RH, Cressman DE, Laz TM, et al. PRL-1, a unique nuclear protein tyrosine phosphatase, affects cell growth. Mol Cell Biol. 1994;14:3752–3762. 66. Zeng Q, Dong JM, Guo K, et al. PRL-3 and PRL-1 promote cell migration, invasion, and metastasis. Cancer Res. 2003 1; 63: 2716–2722.
67. Mark J, Dahlenfors R, Ekedahl C. On double-minutes and their origin in a benign human neoplasm, a mixed salivary gland tumour. Anticancer Res. 1982;2:261–264. 68. Hall JG. Genomic imprinting: Review and relevance to human diseases. Am J Hum Genet. 1990;46:857–873. 69. Heutink P, van Der Mey AG, Sandkuijl LA, et al. A gene subject to genomic imprinting and responsible for hereditary paragangliomas maps to chromosome 11q23-qter. Hum Mol Genet. 1992;1:7–10. 70. Mariman EC, van Beersum SE, Cremers CW, et al. Analysis of a second family with hereditary non-chromaffin paragangliomas locates the underlying gene at the proximal region of chromosome 11q. Hum Genet. 1993;91:357–361. 71. Baysal BE, Ferrell RE, Willett-Brozick JE, et al. Mutations in SDHD, a mitochondrial complex II gene, in hereditary paraganglioma. Science. 2000;287:848–851. 72. Baysal BE, Farr JE, Rubinstein WS, et al. Fine mapping of an imprinted gene for familial nonchromaffin paragangliomas, on chromosome 11q23. Am J Hum Genet. 1997;60:121–32. 73. Niemann S, Steinberger D, Muller U. PGL3, a third, not maternally imprinted locus in autosomal dominant paraganglioma. Neurogenetics. 1999;2:167–170. 74. Baysal BE. On the association of succinate dehydrogenase mutations with hereditary paraganglioma. Trends Endocrinol Metab. 2003;14:453–459. 75. Bikhazi PH, Messina L, Mhatre AN, et al. Molecular pathogenesis in sporadic head and neck paraganglioma. Laryngoscope. 2000;110:1346–1348. 76. Astrom K, Cohen JE, Willett-Brozick JE, et al. Altitude is a phenotypic modifier in hereditary paraganglioma type 1: Evidence for an oxygen-sensing defect. Hum Genet. 2003;113:228–237. 77. Bikhazi PH, Roeder E, Attaie A, et al. Familial paragangliomas: The emerging impact of molecular genetics on evaluation and management. Am J Otol. 1999;20:639–643. 78. Stephenson CF, Bridge JA, Sandberg AA. Cytogenetic and pathologic aspects of Ewing’s sarcoma and neuroectodermal tumors. Hum Pathol. 1992;23:1270–1277. 79. Gil Z, Constantini S, Spektor S, et al. Skull base approaches in the pediatric population. Head Neck. 2005;27:682–689. 80. Barr FG, Galili N, Holick J, et al. Rearrangement of the PAX3 paired box gene in the paediatric solid tumour alveolar rhabdomyosarcoma. Nat Genet. 1993;3:113–117. 81. Davis RJ, D’Cruz CM, Lovell MA, et al. Fusion of PAX7 to FKHR by the variant t(1;13)(p36;q14) translocation in alveolar rhabdomyosarcoma. Cancer Res. 1994;54:2869–2872. 82. Galili N, Davis RJ, Fredericks WJ, et al. Fusion of a fork head domain gene to PAX3 in the solid tumour alveolar rhabdomyosarcoma. Nat Genet. 1993;5:230–235. 83. Gordon T, McManus A, Anderson J, et al.; United Kingdom Children’s Cancer Study Group; United Kingdom Cancer Cytogenetics Group. Cytogenetic abnormalities in 42 rhabdomyosarcoma: A United Kingdom Cancer Cytogenetics Group Study. Med Pediatr Oncol. 2001;36:259–267. 84. Kullendorff CM, Donner M, Mertens F, et al. Chromosomal aberrations in a consecutive series of childhood rhabdomyosarcoma. Med Pediatr Oncol. 1998;30:156–159. 85. Limon J, Dal Cin P, Sandberg AA. Translocations involving the X chromosome in solid tumors: Presentation of two sarcomas with t(X;18)(q13;p11). Cancer Genet Cytogenet. 1986;23 (1): 87–91. 86. Turc-Carel C, Dal Cin P, Limon J, et al. Involvement of chromosome X in primary cytogenetic change in human neoplasia: Nonrandom translocation in synovial sarcoma. Proc Natl Acad Sci USA. 1987;84:1981–1985. 87. Sandberg AA, Bridge JA. Updates on the cytogenetics and molecular genetics of bone and soft tissue tumors. Synovial sarcoma. Cancer Genet Cytogenet. 2002;133:1–23. 88. O’Sullivan MJ, Kyriakos M, Zhu X, et al. Malignant peripheral nerve sheath tumors with t(X;18). A pathologic and molecular genetic study. Mod Pathol. 2000;13:1336–1346. 89. Clark J, Rocques PJ, Crew AJ, et al. Identification of novel genes, SYT and SSX, involved in the t(X;18)(p11.2;q11.2) translocation
Chapter 3: Genetic Abnormalities of Skull Base Tumors
90.
91.
92.
93.
94.
95.
96.
found in human synovial sarcoma. Nat Genet. 1994;7:502– 508. Shipley JM, Clark J, Crew AJ, et al. The t(X;18)(p11.2;q11.2) translocation found in human synovial sarcomas involves two distinct loci on the X chromosome. Oncogene. 1994;9:1447–1453. Brett D, Whitehouse S, Antonson P, et al. The SYT protein involved in the t(X;18) synovial sarcoma translocation is a transcriptional activator localised in nuclear bodies. Hum Mol Genet. 1997;6:1559–1564. Inagaki H, Nagasaka T, Otsuka T, et al. Association of SYT-SSX fusion types with proliferative activity and prognosis in synovial sarcoma. Mod Pathol. 2000;13:482–488. Mandahl N, Heim S, Willen H, et al. Characteristic karyotypic anomalies identify subtypes of malignant fibrous histiocytoma. Genes Chromosomes Cancer. 1989;1:9–14. Rydholm A, Mandahl N, Heim S, et al. Malignant fibrous histiocytomas with a 19p+ marker chromosome have increased relapse rate. Genes Chromosomes Cancer. 1990;24:296–299. Choong PF, Mandahl N, Mertens F, et al. 19p+ marker chromosome correlates with relapse in malignant fibrous histiocytoma. Genes Chromosomes Cancer. 1996;16:88–93. Sandberg AA, Bridge JA. Updates on the cytogenetics and molecular genetics of bone and soft tissue tumors: Osteosar-
97.
98.
99.
100.
101.
102. 103.
79
coma and related tumors. Cancer Genet Cytogenet. 2003;145:1– 30. Argani P, Perez-Ordonez B, Xiao H, et al. Olfactory neuroblastoma is not related to the Ewing family of tumors: Absence of EWS/FLI1 gene fusion and MIC2 expression. Am J Surg Pathol. 1998;22:391–8. Patel NP, Mhatre AN, Lalwani AK. Molecular pathogenesis of skull base tumors. Otol Neurotol. 2004;25:636– 643. Kawai A, Woodruff J, Healey JH, et al. SYT-SSX gene fusion as a determinant of morphology and prognosis in synovial sarcoma. N Engl J Med. 1998;338:153–160. Barr FG, Chatten J, D’Cruz CM, et al. Molecular assays for chromosomal translocations in the diagnosis of pediatric soft tissue sarcomas. JAMA. 1995;273:553–557. Naito N, Kawai A, Ouchida M, et al. A reverse transcriptasepolymerase chain reaction assay in the diagnosis of soft tissue sarcomas. Cancer. 2000;89:1992–1998. Skubitz KM, Skubitz AP. Characterization of sarcomas by means of gene expression. J Lab Clin Med. 2004;144 (2):78–91. Gil Z, Orr-Urtreger A, Voskoboinik N, et al. Cytogenetic analysis of sinonasal carcinomas. Otolaryngol Head Neck Surg. 2006;134 (4):654–60.
4 Imaging of Skull Base Neoplasms Lawrence E. Ginsberg
lesions, such as glomus tumors or juvenile angiofibroma, to establish that a lesion is indeed hypervascular and then to assist in surgical planning. The advantages of MR include its ability to distinguish tumor within a paranasal sinus from obstructed secretions, something not always possible with CT. MR is better able to detect small soft tissue tumor components, particularly those near bony surfaces for which the enhancement may be inconspicuous on CT. MR is also more accurate to detect tumor extension through neural foramina and canals, whether by direct or perineural mechanisms.
INTRODUCTION Management of skull base neoplasms requires a team approach, whose members are dedicated and specifically experienced at dealing with these unique tumors. The radiologist is no exception. Imaging plays a vital role in the diagnosis and posttreatment evaluation of skull base tumors. Although radiologists traditionally concern themselves with preoperative diagnosis, it will be my argument that because such a goal is not attainable in every case, there are more important things for us to achieve with our time and technology. The goal of imaging of known or suspected skull base tumors is several-fold. Firstly, some patients not known but only suspected of having a possible skull base tumor (based on pain, cranial neuropathy, etc.) require imaging simply to establish or exclude such a diagnosis. For those with known skull base tumors, the role of imaging is to establish the full extent and location of the abnormality, outline areas of possible spread and secondary effects on adjacent structures, exclude nodal disease, and finally, suggest possible histologic diagnosis (1). In some cases, distinguishing tumor from benign disease entities such as skull base osteomyelitis and other inflammations is a critical role of imaging. Obviously when possible, the radiologist can suggest a diagnosis, and in fact for many lesions the imaging is quite characteristic (1,2). The goals of this chapter will be to provide an overview of the imaging appearances of the more common skull base neoplasms, provide imaging strategies, and review ways in which certain imaging features provide clues as to tumor type or benign versus malignant. It will not be the goal to exhaustively depict every conceivable skull base tumor.
ANTERIOR CRANIAL BASE The anterior cranial base comprises the orbital and ethmoid roofs, and the cribriform plates. Tumors seldom primarily arise within these bony structures—most generally originate intracranially and extend inferiorly through the skull base (3) [e.g., meningioma (Fig. 1)], or by extending intracranially from an origin in the upper nasoethmoid or frontal sinus region. The latter are typically sinonasal malignancies (Figs. 2 and 3). Intracranial meningiomas are quite common, and often arise in the subfrontal or olfactory groove region (Fig. 1) (4). These benign dural tumors may extend inferiorly, growing directly through the ethmoid roof (fovea ethmoidalis) and cribriform plate into the nasal cavity and/or ethmoid sinuses (Fig. 1). In such cases, it is generally evident radiographically that the bulk (or so-called “epicenter”) of the tumor is intracranial. In addition, characteristic imaging features of meningioma, such as relatively homogeneous isointensity on T1 , iso or slightly hyperintense signal on T2 -weighted images, and bright homogeneous enhancement following Gadolinium-based IV contrast administration, often with a so-called dural tail of enhancement, make the diagnosis of meningioma straightforward in most cases (Fig. 1) (4). Sinonasal malignancies may arise in the upper nasoethmoid region, and typical histologic tumor types include olfactory neuroblastoma (esthesioneuroblastoma), sinonasal undifferentiated carcinoma, squamous cell carcinoma, and neuroendocrine carcinoma (3). Other tumor types are less common. While the imaging characteristics of these lesions are relatively nonspecific, imaging is important in evaluating for intracranial spread, which will often have implications for surgical therapy. CT can often make this determination, but MR is better in this regard (Figs. 2 and 3). On CT, sinonasal malignancies generally enhance to a mild or moderate degree, and when they involve bone, the pattern is usually one of destruction; sclerosis of bone is uncommon. With MR, sinonasal malignancies are typically isointense (same signal as) on T1 - and T2 -weighted sequences, and enhance to
SELECTION OF IMAGING MODALITY As opposed to some anatomic sites in the head and neck, for lesions of the sinonasal cavity and skull base, MR and CT are complimentary. At the M. D. Anderson Cancer Center, virtually all patients with such lesions will be imaged with both modalities prior to therapy, and many will have both modalities following therapy, at least early on. The advantages of CT include its ability to detect calcification and bone, which is important in lesions that either destroy bone or produce some characteristic bony or calcific change. The latter include the classic sunburst periosteal new bone formation in osteosarcoma, the mineralized chondroid matrix of chondrosarcoma, or the hyperostotic reaction typical of certain types of meningioma, among others. CT may also have value, due to the multiplanar reconstruction capability of multidetector-row technology, to provide images in virtually any plane of section, of very high quality. Furthermore, CT angiography has value in certain hypervascular 81
82
Ginsberg
(A)
(B)
(C)
(D)
Figure 1 Olfactory groove meningioma in a 53-year-old woman complaining of visual disturbance. (A–C) Axial T1 -weighted, T2 -weighted, and postcontrast T1 -weighted axial MR images, respectively, demonstrate a large subfrontal mass (asterisks). This has classic signal characteristics and bright homogenous enhancement typical of meningioma. The T2 -weighted image demonstrates so-called “CSF-clefts,” representing pockets of trapped CSF between the lesion and brain surface that is characteristic of an extra-axial lesion (arrows in B). (D) Coronal postcontrast T1 -weighted image demonstrating downward, nasoethmoid tumor extension (arrows).
varying degrees with contrast administration (Figs. 2 and 3). Obstructed sinonasal secretions are usually low signal on T1 and high signal on T2 , and if chronic or inspissated, may become hyperintense (high signal) on T1 -weighted images (Fig. 3) (3). By reviewing all sequences, one can generally determine whether sinus is involved with tumor or merely obstructed. It is important to always image the neck in cases of sinonasal malignancy, as such lesions may present with, or recur as, nodal disease.
CENTRAL SKULL BASE The central skull base (CSB) includes primarily the sphenoid bone and its various parts as well as adjacent structures such as the cavernous sinus, sella turcica, and parasellar region. A very large variety of tumors may afflict the CSB (1–3). As with the anterior cranial base, lesions may arise intracranially and secondarily involve the sphenoid bone, arise primarily within
(A)
(B)
Figure 2 Squamous cell carcinoma of the frontal and ethmoid sinuses in a 66-year-old woman presenting with bloody nasal and oral secretions and subsequent left eye visual changes. (A and B) Coronal and sagittal T1 weighted postcontrast MR images, respectively, demonstrate a large, somewhat heterogeneously but brightly enhancing mass lesion filling the frontal and nasoethmoid regions with obvious posterior extension into the epidural space (arrows).
it, or arise inferiorly or anteriorly, and secondarily affect it by upward or posterior spread. The most common intracranially arising lesion to affect the CSB is meningioma (4). Various aspects of the CSB may be involved, including the planum sphenoidale and tuberculum sella, the anterior clinoid, the greater sphenoid wing, the parasellar region and cavernous sinus, and the petroclival region (Figs. 4 and 5). In most of these locations, the lesion is merely dural in location; while there may be some reactive bony sclerosis or hyperostosis, the bone is uninvolved (4). Imaging features of meningioma in these locations are no different than elsewhere. There is one exception, hyperostosing en plaque meningioma of the greater sphenoid wing, which is associated with a very striking sclerotic response, or hyperostosis (4,5). This bony reaction indicates actual through and through tumor involvement of the bone, and has a very characteristic radiographic appearance, including intraorbital extension (Fig. 6) (1,4,5). While most meningiomas remain within only the dura, in addition to the hyperostosing en plaque meningioma of the greater sphenoid wing, some meningiomas may invade the CSB in an aggressive, almost malignant manner (1,4). The imaging for such invasion can be fairly dramatic, and if the diagnosis of meningioma is unknown, the radiologic diagnosis may be more difficult (Fig. 7). Such meningiomas may achieve massive size, or extend directly through the bone, or through neural foramina, into extracranial spaces such as the pterygopalatine fossa, orbit, or masticator space. Such lesions may be grossly destructive of bone (Fig. 7). It should be kept in mind, for any location, that dural metastases may be radiographically indistinguishable from meningioma. In a known cancer patient, what may resemble on a first imaging study a meningioma may be a dural metastasis, and the finding of growth on serial imaging should raise that possibility (Fig. 8) (6,7). Finally, another lesion that might be radiographically confused with meningioma is the rare, highly vascular malignancy, hemangiopericytoma. These also have the potential to be highly destructive of the cranial base (Fig. 9). Another common benign CSB lesion is the pituitary adenoma (1,8–11). Some macroadenomas (those greater than 1 cm in size), often but not always endocrinologically nonfunctioning, may achieve a very large size, and be invasive of the central skull base. Such cases may present a diagnostic
Chapter 4: Imaging of Skull Base Neoplasms
(A)
(B)
83
(C)
(D)
(E)
Figure 3 Olfactory neuroblastoma (esthesioneuroblastoma) in the upper nasal cavity in a 47-year-old man complaining of anosmia and subsequent development epistaxis and congestion. (A) Axial postcontrast CT demonstrating relatively nondescript and nonenhancing soft tissue filling left nasoethmoid region and to a lesser extent the right nasal cavity (arrows). There is an enlarged, airless left ethmoid air cell mucocele that is slightly less dense than the tumor (arrowhead). Notice that there is no overt bone destruction evident, though there could be in this entity. (B–D) Axial T1 -weighted, T2 -weighted, and postcontrast axial T1 -weighted MR images, respectively, demonstrate the characteristic findings of sinonasal malignancy. The lesion is isointense to muscle and other soft tissue (arrows in B). Notice on the T2 -weighted image (arrows in C), the tumor is relatively isointense to brain and not nearly as high signal or hyperintense as obstructed secretions in the left ethmoid mucocele (arrowheads in C). (D) The lesion enhances brightly, whereas obstructed secretions in the mucocele do not (arrowhead in D). (E) Coronal postcontrast T1 -weighted MR image shows intracranial extension through the cribriform plate and ethmoid roof (arrows).
(A)
(B)
(C)
Figure 4 53-year-old woman presenting with decreased visual acuity in the right eye. Dx: optic sheath/canal meningioma with involvement of the clinoid region and planum sphenoidale. (A–C) Axial, coronal, and sagittal T1 -weighted postcontrast MR images, respectively, demonstrate enhancing tumor in a characteristic elongated manner along the surface of the right optic nerve (arrows in A). Tumor extends up onto the planum as can be seen by arrows in B and C.
84
Ginsberg
(A)
(B)
Figure 5 Noninvasive greater sphenoid wing meningioma in an asymptomatic 75-year-old woman. This lesion was picked up on a routine lymphoma screening CT. (A and B) Axial and coronal postcontrast T1 -weighted MR images, respectively, reveal a homogenously brightly enhancing mass lesion (arrows) along the left parasellar and greater sphenoid wing region. There is an obvious dural tail or enhancement of dura extending away from the tumor, characteristic but not diagnostic of meningioma (arrowhead in A).
challenge to the radiologist, and there is no specific imaging feature, other than a tendency for the lesion to encompass the sella. It is important, therefore, for the radiologist to consider pituitary adenoma when confronted with a large, destructive central skull base mass, and to at least check hormone levels, specifically prolactin, in such cases (Fig. 10). One relatively uncommon benign tumor that arises near and may secondarily affect the CSB is the juvenile angiofibroma. These highly vascular tumors typically present with epistaxis in teenage males and have very characteristic imaging features (1,12). They arise in or near the sphenopalatine foramen, the opening between the pterygopalatine fossa and the nasal cavity. There are often tumor components in the nasal cavity and nasopharynx, and these aggressive lesions often invade into the clivus and sphenoid sinus. Because of the hypervascularity, these lesions enhance brightly on CT, and have small foci of low signal, so-called flow voids on MR, representing rapidly flowing arterial supply to the tumor (Fig. 11). Vascular studies such as catheter angiography, CT angiography, or MR angiography have value in providing
(A)
(B)
a surgical road map or to facilitating preoperative embolization [Fig. 11(E)]. One final intracranial (sometimes both intra- and extracranial) lesion to affect the CSB is the neural sheath tumors arising from the trigeminal nerve, schwannoma, and neurofibroma (2,6). These usually benign lesions can arise from the main trigeminal trunk and extend variously along its branches or may arise within one of its three divisions. Such lesions often cause widening of foramina or other benign bony remodeling (rather than destruction) of the central skull base (Fig. 12). Among malignancies that may affect the CSB, which arise primarily extracranially, probably the most common is nasopharyngeal carcinoma (NPC) (1,8,13,14). Given its location immediately inferior to the clivus, upward spread of NPC very commonly involves the CSB. In some cases, the nasopharyngeal component may be relatively small (Fig. 13), and in such cases the radiologist must bear in mind the possibility of NPC. As opposed to some malignancies that tend to destroy bone, and despite the fact that NPC may do this, it often has a tendency to infiltrate in a nondestructive manner (15). For this reason, MR imaging is more sensitive than CT for detecting such involvement (Fig. 14). Of course, NPC can also cause skull base destruction, in which CT has a role as well (15). Because MR is also more sensitive for detecting intracranial spread via perineural and direct mechanisms, it is our preference to use MR as the main imaging modality in cases of NPC (16). Another type of extracranial malignancy that may secondarily affect the CSB is head and neck malignancy with perineural tumor spread (PNS) (17–19). This is a very extensive topic, and a thorough description is beyond the scope of this chapter. The most common manner in which the CSB is affected is in a patient with a known, or sometimes unknown, malignancy of the face (most commonly in a V2 distribution, like the cheek) (Fig. 15) or a lesion of the lower lip or palate (or other less common mucosal surface). Common tumor types include squamous cell carcinoma, adenoid cystic carcinoma, and desmoplastic melanoma (18). Such lesions may spread proximally, toward the CNS, along the branch of the trigeminal nerve that innervates the primary site. For the cheek and palate, this means the infraorbital (Fig. 15) or palatine nerves (Fig. 16), respectively, with tumor spread to the pterygopalatine fossa (Figs. 15 and 16). From here, tumor can spread
(C)
Figure 6 Hyperostosing en plaque greater sphenoid wing meningioma in a 54-year-old woman presenting with left-sided headaches and progressive visual loss in the left eye. (A) Axial bone window CT image demonstrates gross bone thickening or hyperostosis of the left greater sphenoid wing and lateral orbital wall (arrows). (B and C) Axial precontrast and fat-suppressed postcontrast T1 -weighted MR images, respectively, show that the bone marrow is markedly abnormal (arrows in B) as opposed to the normal marrow in the right sphenoid triangle depicted with an arrowhead. Enhancing soft tissue (extraosseous) tumor can be seen in the left orbit and behind the greater sphenoid wing in C (arrows).
Chapter 4: Imaging of Skull Base Neoplasms
(A)
(B)
(D)
85
(C)
(E)
Figure 7 Massive skull base meningioma in a 61-year-old woman presenting with progressive vision loss in the left eye and subsequent development of proptosis. (A) Axial CT bone window demonstrates a permeative type of bone destruction (arrows). (B–D) Axial T1 -weighted, T2 -weighted, and postcontrast axial T1 -weighted MR images, respectively, demonstrate extensive soft tissue tumor infiltrating the nasoethmoid region, left cavernous sinus, and left orbit (arrows). Note the dural tail along the incisura and sphenoid wing (arrowheads in D). (E) Coronal postcontrast MR image demonstrates the extensive nature of this lesion.
posteriorly along the main trunk of the maxillary nerve, through foramen rotundum, and into the cavernous sinus or further posteriorly into Meckel cave or even onto the main trigeminal trunk (18). For lesions of the lower lip or gingiva, tumor can access the inferior alveolar branch of V3 , and then upward onto the main trunk of the mandibular nerve
(Fig. 17), through foramen ovale, and into Meckel’s cave. Though perhaps best considered in the temporal bone section, the facial nerve may also be involved with tumor spread, mainly from tumors originating within, or that spread to, the parotid gland (Fig. 18). Recognition of this type of spread is a critical function of the radiologist because failure to recognize,
(A)
Figure 9 Masticator space and middle cranial fossa/skull base hemangiopericytoma in a 45-year-old man who presented with headache, nausea, and vomiting. Coronal postcontrast T1 -weighted MR image demonstrates a large, moderately enhancing mass encompassing the middle cranial fossa and obvious destruction through the calvarium (arrows). There is a central area of nonenhancement, representing necrosis. This is indistinguishable from an aggressive meningioma but proved to be a hemangiopericytoma at surgery.
(B)
Figure 8 Dural-based planum sphenoidale metastasis initially believed to be meningioma in a 47-year-old woman with metastatic breast carcinoma. (A) Sagittal T1 postcontrast MR image demonstrates a small enhancing lesion along the planum sphenoidale (arrow). Note similarity with the meningioma shown in Figure 4. (B) Sagittal T1 -weighted postcontrast MR image 8 months later shows clear progression of this lesion (arrows), which prompted surgery.
86
Ginsberg
(A)
(B)
(C)
(D)
Figure 10 Massive pituitary adenoma in a 48-year-old man who presented with only severe headache. (A) Axial CT bone window reveals a large destructive process involving the central skull base (arrows). (B and C) Axial T2 and postcontrast T1 -weighted MR images, respectively, reveal a large minimally and somewhat heterogeneously hyperintense and brightly enhancing mass involving virtually the entire central skull base. (D) Sagittal image revealing complete replacement of the sphenoid sinus and clivus. Notice while the lesion encompasses the sella (asterisk), there is no suprasellar extension, and thus the diagnosis of pituitary adenoma less than obvious. In this patient, the prolactin level was 344214, indicating prolactinoma.
(A)
(B)
(C)
(D)
(E)
Figure 11 Juvenile angiofibroma in a 20-year-old man presenting with bilateral nasal congestion but no epistaxis. (A) Axial CT bone window revealing a large mass with destruction of the mid and left aspect of the central skull base (arrows). (B) Contrast-enhanced CT soft tissue window reveals a moderately enhancing mass lesion corresponding to the areas of bone destruction (arrows). (C and D) Axial T2 - and T1 -weighted postcontrast MR images, respectively, reveal a hyperintense, enhancing lesion. Central areas of signal void represent blood vessels and are so-called “flow voids” (arrows in C). (E) Lateral view of an external carotid angiogram in a different patient with a juvenile angiofibroma reveals marked tumor hypervascularity.
(A)
(B)
(C)
(D)
Figure 12 Massive trigeminal/skull base schwannoma in a 44-year-old man who presented with right eye swelling and intermittent visual complaints. (A) Axial CT bone window reveals benign bony remodeling (arrows) with the lateral wall the sphenoid sinus pushed medially. (B–D) Axial T2 postcontrast, axial, and coronal postcontrast T1 MR images, respectively, reveal a very large heterogeneously T2 hyperintense and brightly enhancing extra-axial mass lesion centered at and above the floor of the right middle cranial fossa. The coronal image shows obvious tumor extending into the sphenoid sinus (arrow) and elevation the right temporal lobe (arrowhead). This was not a straightforward diagnosis, but the lack of bony destruction or edema in the brain argued for a slow growing process such as, in this case, schwannoma.
Chapter 4: Imaging of Skull Base Neoplasms
(A)
(B)
(C)
87
(D)
Figure 13 Nasopharyngeal carcinoma with massive central skull base destruction in a 55-year-old man presenting with headache and subsequent development of diplopia. (A) Axial CT bone window reveals gross destruction of the central skull base (arrows). (B) Axial T1 postcontrast MR image demonstrates a small mass in the right fossa of Rosenm¨uller (arrow). There is also abnormal enhancement in the clivus, indicating tumor involvement (arrowhead). (C and D) Axial T1 and T1 postcontrast MR images, respectively, show replacement of the normal T1 hyperintense bone marrow of the clivus (asterisks in C), and heterogeneously bright contrast enhancement. Skull base involvement in NPC indicates T3 staging.
very common unfortunately, can have a very adverse effect on treatment outcome. Commonly, PNS occurs in the form of, or at the time of, tumor recurrence (Fig. 15), not necessarily at the time of diagnosis or initial therapy. In other cases, sadly, PNS represents progression of disease that was unrecognized initially, and thus not addressed clinically. In terms of malignancies arising in the skull base directly, metastases are the most common (3). Such lesions may be lytic or blastic, depending on the primary malignancy, but both are common (Fig. 19). The imaging is generally straightforward, except to say that small lesions may be subtle. Primary CSB malignancies include chordoma and chondrosarcoma. Chordomas are notochord-derived malignancies that occur in the sacrococcygeal region (50%), spine (15%), and clivus (35%) (1,3). They may also occur off midline occasionally. Most chordomas occur in mid-late adulthood.
Radiographically, they appear as expansile lytic lesions, with internal areas of fragmented bone that are most readily seen on CT (Fig. 20), and on MR they have characteristic T2 signal hyperintensity (Fig. 20). A soft tissue mass is usually present, and some degree of extracranial and/or intracranial extension is typical, the latter often resulting in brain stem compression. Chondrosarcomas may also arise in the CSB (1,3). These cartilage-derived malignancies tend to occur in proximity to the various synchondroses. Common locations include the petroclival fissure and the upper nasal septum/vomer region (Figs. 21–23). They may also involve the greater sphenoid wing more laterally, often with masticator space components (Fig. 23). In terms of imaging appearance, many, though not all, will have a mineralized chondroid matrix seen in chondroid tumors anywhere—a grouping of seemingly organized calcific elements (Figs. 21 and 23). On MR, though the signal intensity varies, there is often marked heterogeneity, particularly on postcontrast T1 -weighted images (Fig. 22). A lesion in the appropriate location and with typical imaging features should suggest the possibility of chondrosarcoma.
POSTERIOR CRANIAL BASE
(A)
(B)
Figure 14 Nasopharyngeal carcinoma. (A) Axial CT bone window shows no discernible abnormality in the lower clivus. (B) Axial T1 -weighted MR image shows abnormal hypointense signal in the right side of the clivus (arrows), indicating tumor infiltration. The primary cancer is also visible (squiggly arrow in B). On the T1 MR image, the marrow should be hyperintense or white in color, as seen in the left side of the clivus and left mandibular condyles (asterisks in B). Also, note metastatic disease in the right parotid region (arrowhead in B), and retained secretions in the right mastoid air cells, owing to Eustachian tube obstruction (arrows in A). This case demonstrates the utility of MRI to evaluate for skull bass infiltration in cases where CT may be normal.
For the purpose of this chapter, the posterior cranial base will include the temporal bone, and the lower occipital structures such as the foramen magnum, hypoglossal canal, and the jugular foramen/carotid canal. There is a large variety of tumors that may affect the temporal bone. In terms of extracranial lesions affecting it secondarily, lesions of the external ear are probably the most common (20). Periauricular lesions may also grow into the temporal bone. Histologic tumor types include squamous cell, adenoid cystic carcinoma of the external canal, melanoma, and basal cell carcinoma (20). When imaging skin lesions in and around the ear, CT with high-resolution bone windows is critical to assess the bony structures (Fig. 24.) The temporal bone may also be infiltrated by petrous region meningiomas. While the diagnosis of meningioma is often fairly straightforward, occasionally a meningioma may be primarily infiltrative, without much dural disease; in such cases the diagnosis can be difficult as the imaging may be
88
Ginsberg
(A)
(B)
(C)
(D)
Figure 15 Recurrent squamous cell carcinoma in the left premaxillary region in a 62-year-old man who had previously undergone Mohs resection, now presenting with left facial (V2 ) paresthesias. This case demonstrates perineural tumor spread along the infraorbital branch of the maxillary nerve. (A) Axial postcontrast CT image demonstrates a subcutaneous mass anterior to the left maxillary sinus (straight arrow). There is also abnormal density within the infraorbital foramen and anterior aspect of the left infraorbital canal (curved arrow). (B and C) Axial pre- and postcontrast T1 -weighted MR images, respectively, reveal enhancing tumor in a premaxillary region (small arrows) as well as abnormal signal and enhancement along the course of the infraorbital nerve (curved arrow). More posteriorly, tumor can be seen actually coursing into the pterygopalatine fossa (large arrow in B). (D) Coronal T1 postcontrast MR with fat suppression reveals enlargement of the infraorbital nerve along the orbital floor or maxillary sinus roof (arrow).
(A)
(B)
(C)
Figure 16 Adenoid cystic carcinoma of the left hard palate with perineural spread along the palatine branches of the maxillary nerve in a 55-year-old whose dentist discovered a palatal mass. There was no clinical evidence of neuropathy. (A) Coronal T1 postcontrast MR image reveals a mass in the left side of the hard palate (arrow). (B) Axial CT bone window shows enlargement of the left greater palatine foramen (arrow). Note the normal right greater palatine foramen (arrowhead). (C) Axial postcontrast T1 MR image shows abnormal enhancement in the left pterygopalatine fossa (arrow), indicating perineural tumor spread to at least this level. From here, tumor can spread posteriorly, in a retrograde fashion, either through foramen rotundum or along the Vidian nerve.
(A)
(B)
(C)
(D)
Figure 17 Recurrent right lower lip squamous cell carcinoma in a 42-year-old man who had previously undergone biopsy and Mohs surgery as well as subsequent wedge resection and neck dissection for prior recurrences. (A) Axial T1 -weighted MR image reveals a large recurrence in the right lower lip, very close to the mental foramen (arrows). Note the probable tumor infiltration of the right mandibular marrow cavity (arrowheads). (B and C) Axial pre- and postcontrast T1 -weighted MR images, respectively, reveal abnormal signal intensity and enhancement in the mandibular foramen for the right inferior alveolar nerve (arrows). (D) Coronal T1 -postcontrast MR image reveals tumor extending in an upward and medial direction from the mandibular foramen along the course of the main mandibular nerve trunk (arrows), approaching foramen ovale (arrowhead).
Chapter 4: Imaging of Skull Base Neoplasms
(A)
(B)
89
(C)
Figure 18 Parotid salivary duct carcinoma associated with perineural spread along the descending segment of the right facial nerve in a 67-year-old man who presented with right-sided facial neuropathy for which an initial MR was said to be negative. (A) Axial noncontrast T1 -weighted MR image reveals a mass in the right parotid gland (arrows). Note the posterior extension of disease toward the stylomastoid foramen (curved arrow). (B and C) axial T1 postcontrast MR images at progressively cephalad positions revealing enlargement and excessive enhancement of the right descending facial nerve segment (arrows in B and C).
(A)
(B)
Figure 19 Two patients with skull base metastasis. (A) Axial CT bone window demonstrating a large lytic and destructive metastasis from lung carcinoma (arrow). (B) Axial CT bone window in a patient with metastatic breast carcinoma showing blastic lesions in the clivus (arrows) and the right side of the central skull base lateral to the Vidian canal (arrowhead).
(A)
(B)
quite nonspecific (Fig. 25). In such cases, only biopsy can differentiate from infiltrative malignancies such as sarcoma or leukemic infiltration (Fig. 26). Intrinsic temporal bone tumors are relatively uncommon, though of course, metastases can occur in the temporal bone (Fig. 27). Vestibular schwannoma will be addressed below in the discussion of cerebellopontine lesions. However, schwannomas of the facial nerve typically present in the intratympanic portion of the facial nerve, and thus deserve mention in a discussion of temporal bone tumors. These benign lesions typically present with facial neuropathy (21). Commonly involved segments include the geniculate ganglion, descending facial nerve, and the greater superficial petrosal nerve. The latter is a branch of the facial nerve containing preganglionic parasympathetic fibers destined to innervate the lacrimal gland and vasomotor nerves of the nasal cavity and palate (22). The nerve emerges from the geniculate ganglion and exits the facial hiatus to become intracranial, along the sphenopetrosal synchondrosis in the floor of the middle cranial fossa (22). A tumor along this portion of the facial nerve can present as an extraaxial mass along the roof, or intracranial surface of the petrous bone, and may be confused radiographically with
(C)
(D)
Figure 20 Clival chordoma in a 40-year-old woman complaining of occipital region headache radiating to the right neck. (A) CT bone window demonstrates a large destructive lesion involving the clivus (arrows). (B–D) Axial T1 , T2 , and postcontrast T1 -weighted MR images, respectively, reveal a well-circumscribed mass lesion in the central skull base that is isointense on T1 , hyperintense on T2 , and enhances moderately with some heterogeneity (arrows). The very bright signal on T2 is very suggestive of chordoma.
90
Ginsberg
(A)
(A)
(B)
Figure 21 Nasoethmoid chondrosarcoma in a 38-year-old woman presenting with progressive sinus symptoms including congestion and epistaxis. (A) Axial CT bone window reveals a destructive mass with internal somewhat organized areas of calcification representing the chondroid matrix (arrows). (B) Axial T2 -weighted MR image revealing a relatively homogeneous isointense mass in the nasoethmoid region (arrows). Chondrosarcomas are often, but not always, heterogeneous on T2 and postcontrast MR images, as in the next figure.
Figure 22 Nasoethmoid chondrosarcoma in a different patient. Axial T1 postcontrast MR image reveals the marked heterogeneity that is characteristic of some chondrosarcomas. Obviously the appearance can be variable.
a meningioma. Recognizing that branches of the facial nerve or enlargement of the facial hiatus, along with the history of facial neuropathy, should suggest the correct diagnosis (Fig. 28).
(A)
(B)
(B)
Figure 24 Adenoid cystic carcinoma of the left external auditory canal in a 60-year-old man with a 2-year history of “ear congestion” and intermittent otorrhea. (A) Axial postcontrast CT (soft-tissue window) demonstrates a large mass about the left external canal and external ear (arrow). (B) Axial CT bone window reveals gross destruction of the temporal bone (arrow).
Intracranial lesions can involve the posterior cranial base by either growing anteroinferiorly or by arising adjacent to, say, the temporal bone. Examples of anteroinferior extension include posterior fossa meningiomas, which have a tendency to grow through neural foramina such as the jugular foramen (Fig. 29) (4). The meningioma may or may not have a bony reaction, often not, and may be difficult to distinguish from nerve sheath tumors. Nerve sheath tumors, of course, grow through neural foramina by their nature and may involve either the nerves within the jugular foramen (Fig. 30) or less commonly the hypoglossal canal (Fig. 31). These lesions cause a characteristic nondestructive widening or expansion of the neural foramina as seen on CT, and have other typical imaging features on CT and MR (Fig. 30). Other benign intracranial neoplasms adjacent to the temporal bone include the cerebellopontine lesions. These comprise vestibular schwannoma, meningioma, and epidermoid cyst. Vestibular schwannomas are quite variable in size and shape, but generally involve the internal auditory canal (IAC) to some extent. Some are entirely intracanalicular (within the internal auditory canal), and others involve primarily in the cistern but extend into the IAC (Figs. 32 and 33). Vestibular schwannomas typically enhance brightly, but may contain areas of cystic change or necrosis, which do not
(C)
(D)
Figure 23 Left central skull base/masticator space chondrosarcoma in a 51-year-old woman presenting with severe headaches. (A) Coronal bone window CT revealing a mass destroying the floor of the middle cranial fossa (arrow). Note the marked chondroid calcifications. (B–D) Axial T1 , T2 , and postcontrast T1 -weighted MR images, respectively, reveal a somewhat heterogeneous mass that is isointense on T1 , hyperintense on T2 , and enhances brightly following gadolinium administration. This is a fairly classic appearance for chondrosarcoma given the heterogeneity as well as chondroid calcifications.
Chapter 4: Imaging of Skull Base Neoplasms
(A)
(B)
91
(C)
Figure 25 Infiltrating neoplasm of the right temporal bone. The diagnosis ultimately proved to be meningioma. (A) Axial CT bone window revealing a destructive, somewhat permeative pattern of bony infiltration in the right temporal bone (arrows). (B and C) Axial pre- and postcontrast T1 -weighted MR images, respectively, show a T1 isointense, mildly enhancing mass replacing the normal temporal bone structures and lower clivus (arrows). The imaging is nonspecific and the differential diagnosis should include other entities such including metastasis (see Fig. 26).
(A)
(B)
Figure 26 Two patients with nonspecific infiltrative masses of the temporal bone. (A) Axial postcontrast, T1 -weighted MR image reveals a patient with leukemic infiltration of the right temporal bone (arrow). (B) Axial postcontrast T1 -weighted MR revealing a large enhancing mass in the left temporal bone (arrow) that proved to be a high-grade sarcoma. Again, the imaging for these lesions as well as that for the meningioma in Figure 25 is nonspecific.
(A)
(B)
Figure 27 Metastasis to the left temporal bone resulting in facial neuropathy in a patient with metastatic carcinoma of the cervix. (A) CT bone window obtained for evaluation of cervical lymphadenopathy, but prior to the onset of facial palsy, showing normal descending left facial nerve canal (arrow). (B) Axial CT bone window following onset of left facial neuropathy. Note the new destructive lytic lesion corresponding to the location of the left descending facial nerve canal (arrow).
(A)
(B)
Figure 28 Facial schwannoma with greater superficial petrosal nerve and middle cranial fossa involvement in a 53-year-old woman with a manyyear history of facial neuropathy and progressive hearing loss. (A) Axial CT bone window reveals enlargement of the facial hiatus for the left greater superficial petrosal nerve (arrow). There is a soft tissue mass at the location for the geniculate ganglion (arrowhead). (B) Axial postcontrast T1 -weighted MR image reveals a homogeneous, brightly enhancing mass lesion with a component in the middle cranial fossa (large arrow), the labyrinthine portion of the facial nerve (small arrow), and the internal auditory canal (arrowhead).
enhance. Of course, bilateral vestibular schwannomas as associated with type II neurofibromatosis (Fig. 33), along with schwannomas of other cranial nerves, and meningiomas (6). The most common jugular foramen neoplasm, of course, is the glomus tumor (2). These so-called paragangliomas are hypervascular, usually benign but aggressive tumors that arise from paraganglia cells in various location in the body, but in the head and neck, typically along the carotid sheath (glomus caroticum or vagale), or for this discussion, the jugular foramen (glomus jugulare). On CT, these cause a typical “moth-eaten” or permeative pattern of bone destruction (Fig. 34) that is quite characteristic (2). They also enhance quite brightly due to their hypervascular nature. On MR, when 2.5 cm or so is exceeded, they will usually have internal flow voids representing their vascularity (Fig. 33). Coronal CT reconstructions are probably the best way to determine if the lesion is extending into the hypotympanum (glomus jugulotympanicum). There is a role for angiographic studies, whether by catheter or with MR/CT techniques, to provide
92
Ginsberg
(A)
(A)
(B)
(B)
Figure 29 Left-sided CP angle meningioma with involvement of the jugular foramen in a 43-year-old woman presenting with tinnitus and progressive left-sided hearing loss. (A and B) Axial and sagittal postcontrast T1 -weighted MR images, respectively, reveal a dural-based extra axial mass lesion in the left CP angle (large arrows). Note extension into the jugular foramen (small arrow in B). There is also a so-called dural tail of enhancement best seen in the axial image extending laterally from the lesion (smaller arrow in A). (C)
Figure 32 Assortment of vestibular schwannomas. (A) Axial postcontrast T1 -weighted MR image through the posterior fossa reveals a small intracanalicular enhancing mass representing a vestibular schwannoma in the right internal auditory canal (arrow). (B) Axial postcontrast T1 -weighted MR image revealing a dumbbell-like mass extending from the CP angle into a widened left internal auditory canal (arrow). There appears to be a small dural tail that is extending posterolaterally, away from the lesion; as such, a dural tail is not specific for meningioma and can be seen in other lesions. (C) Axial postcontrast T1 -weighted MR image shows a larger vestibular schwannoma that is primarily cisternal in location with only minimal extension into the IAC (arrow). Acoustic vestibular schwannomas come in different sizes and shapes. (A)
(B)
Figure 30 Jugular schwannoma in a 52-year-old man presenting with a hissing sound in the left ear and left-side sensorineural hearing loss. (A) Axial CT bone window reveals benign expansion and remodeling, without destruction, in the left jugular foramen (arrow). (B) Sagittal postcontrast T1 -weighted MR image shows a heterogeneously but brightly enhancing mass lesion extending from the posterior fossa through a widened jugular foramen (arrows) and into the upper carotid space.
(A)
(B)
Figure 31 Left-sided hypoglossal schwannoma in a patient presenting with headache and left tongue fasciculations. (A) Axial postcontrast T1 weighted MR image through the posterior fossa reveals a dumbbell-shaped enhancing tumor with a component in the cerebellomedullary angle (white arrows) and extending out through the hypoglossal canal into the upper carotid space (black arrow in B).
Figure 33 Neurofibromatosis type II in a patient with long-standing disease and multiple manifestations. Axial T1 postcontrast MR image through the posterior fossa and skull base reveals bilateral acoustic vestibular schwannomas, with a larger right-sided lesion (arrowhead) and the left-sided lesion intracanalicular (small arrow). Incidental note is also made of an enhancing mass within the right cavernous sinus (curved arrow) representing a trigeminal schwannoma, and only partially seen on this image, a large plexiform neurofibroma involving the upper eyelid (squiggly arrow).
Chapter 4: Imaging of Skull Base Neoplasms
(A)
(B)
(C)
93
(D)
Figure 34 Massive skull base glomus tumor in a 32-year-old woman with long-standing occipital region headache and progressive lower cranial neuropathies. (A) Axial CT bone window reveals a permeative type of bony destruction in the right temporal bone in the region of the jugular foramen (arrow). (B and C) Axial T2 and sagittal T1 postcontrast MR images, respectively, reveal a large mass lesion. Note internal areas of heterogeneity, some of which are dot like and some of which are curvilinear (arrows) representing flow voids within this hypervascular lesion. The sagittal image nicely demonstrates inferior extension into the upper carotid space. (D) Axial CT bone window through the temporal bone in a different patient with a jugular foramen glomus tumor (black arrows) with extension into the tympanic cavity (white arrows). As such this is properly termed a glomus jugulotympanicum.
vascular road mapping preoperatively or to facilitate embolization.
SUMMARY Tumors of the skull encompass a huge variety of lesions, and even for a given diagnosis, the imaging appearance can vary. Because of this fact and the often nonspecific nature of imaging findings, diagnosis is not always possible prior to biopsy. It is the radiologist’s job to confirm the presence of disease and to define the extent of tumor with precision. For many cases, this will suggest a likely diagnosis. It is hoped through this chapter that the reader has gained an appreciation and healthy respect for the spectrum of imaging findings associated with skull base neoplasms. The skull base radiologist must accept that they will not always be correct, and only through experience can one learn the many nuances of this field. As difficult as initial diagnosis is, posttreatment and especially postoperative imaging in the patient population is even more difficult. There is no substitute for experience and, frankly, the wisdom born of error. REFERENCES 1. Ginsberg LE. Neoplastic diseases affecting the central skull base: CT and MR imaging. Am J Roentgenol. 1992;159:581–589. 2. Harnsberger HR, Wiggins RH, Hudgins PA, et al. Diagnostic Imaging. Head and Neck, 1st ed. Salt Lake City, UT: Amirsys, 2004:1 v. 3. Som PM, Brandwein MS. Tumors and tumor-like conditions. In: Som PM, Curtin HD, eds. Head and Neck Imaging. St. Louis: Mosby, 2003. 4. Ginsberg LE, Moody DM. Meningiomas: Imaging. In: H. WR, S. RS, eds. Neurosurgery. New York: McGraw-Hill, 1996:855–872. 5. Kim KS, Rogers LF, Goldblatt D. CT features of hyperostosing meningioma en plaque. Am J Roentgenol. 1987;149:1017– 1023. 6. Ginsberg LE. Contrast enhancement in meningeal and extra-axial disease. Neuroimaging Clin N Am. 1994;4:133–152.
7. Laidlaw JD, Kumar A, Chan A. Dural metastases mimicking meningioma. Case report and review of the literature. J Clin Neurosci. 2004;11:780–783. 8. Curtin HD, Rabinov J, Som PM. Skull base: Embryology, anatomy, and pathology. In: Som PM, Curtin HD, eds. Head and Neck Imaging. St. Louis: Mosby, 2003. 9. Fischbein NJ, Kaplan MJ. Magnetic resonance imaging of the central skull base. Top Magn Reson Imag. 1999;10:325–346. 10. Levy RA, Quint DJ. Giant pituitary adenoma with unusual orbital and skull base extension. Am J Roentgenol. 1998;170:194–196. 11. Minniti G, Jaffrain-Rea M-L, Santoro A, et al. Giant prolactinomas presenting as skull base tumors. Surg Neurol. 2002;57:99–103. 12. Weinstein MA, Levine H, Duchesneau PM, Tucker HM. Diagnosis of juvenile angiofibroma by computed tomography. Radiology. 1978;126:703–705. 13. Ishida H, Mohri M, Amatsu M. Invasion of the skull base by carcinomas: Histopathologically evidenced findings with CT and MRI. Eur Arch Otorhinolaryngol. 2002;259:535–539. 14. Chong VF, Mukherji SK, Ng SH, et al. Nasopharyngeal carcinoma: Review of how imaging affects staging. J Comput Assist Tomogr. 1999;23:984–993. 15. Chong VF, Fan YF. Skull base erosion in nasopharyngeal carcinoma: Detection by CT and MRI. Clin Radiol. 1996;51:625–631. 16. Chong VF, Fan YF, Khoo JB. Nasopharyngeal carcinoma with intracranial spread: CT and MR characteristics. J Comput Assist Tomogr. 1996;20:563–569. 17. Ginsberg LE, Demonte F. Palatal adenoid cystic carcinoma presenting as perineural spread to the cavernous sinus. Skull Base Surg. 1998;8:39–43. 18. Ginsberg LE. Imaging of perineural tumor spread in head and neck cancer. In: Som PM, Curtin HD, eds. Head and Neck Imaging. St. Louis: Mosby, 2003. 19. Ginsberg LE. Imaging of perineural tumor spread in head and neck cancer. Semin Ultrasound CT MR. 1999;20:175–186. 20. Dinehart SM, Jansen GT. Cancer of the skin. In: Myers EN, Suen JY, eds. Cancer of the Head and Neck. Philadelphia, PA: W.B. Saunders, 1996:143–159. 21. Ginsberg LE, DeMonte F. Diagnosis please. Case 16: Facial nerve schwannoma with middle cranial fossa involvement. Radiology. 1999;213: 364–368. 22. Ginsberg LE, De Monte F, Gillenwater AM. Greater superficial petrosal nerve: Anatomy and MR findings in perineural tumor spread. Am J Neuroradiol. 1996;17:389–393.
5 Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base: Clinical Examination, Auditory Testing, Vestibular Testing, and Equilibrium Paul W. Gidley
complaint of ear pain, and only after a careful history and physical examination does one discover the cause: vocal fold malignancy.
INTRODUCTION The art of history taking, physical examination, and creation of a differential diagnosis develops over one’s practice. Key to this art is the understanding of the subtleties of symptoms and signs, especially with neurologic diseases. This chapter will discuss the basics of the neuro-otologic examination. Sitespecific symptoms and signs will be explored in order to help the physician learn the cause of these maladies. This chapter is meant to be a foundation from which one can work toward a proper differential diagnosis and treatment plan for each patient. The interaction between physician and patient establishes the basic rapport and tenor of the physician-patient relationship. This interaction is supplanted and not replaced by technology. The daily and long-term outcomes from treatment are readily measured and tracked by noting symptoms and signs of disease. This is true whether at the bedside or in the outpatient setting. It is assumed that the reader, being either a resident or well-seasoned practitioner, is already familiar with the basics of history and physical examination. This chapter’s objectives are to discuss the physical signs associated with neuro-otologic disease and to discuss the ancillary audiometric, vestibular, and electrophysiologic tests used to discern disease processes. Special attention is paid to the cranial nerve examination. This chapter will not discuss individual disease states except to discuss the signs and symptoms that are indicative of a particular disease process.
Head, Scalp, and Skin Occasionally, the obvious does escape our attention. The head and neck are covered by skin, but occasionally this skin is overlooked to examine instead the ear canals or nasal passages. The sun-exposed portions need to be examined for premalignant or malignant conditions. Patients who have already had skin excisions need to be queried regarding the pathologic diagnosis from these sites given the propensity of melanoma and squamous cell carcinoma to metastasize or have perineural spread. At times, the skin of the head and neck provides the diagnosis, as in the case of adenoma sebaceum and tuberous sclerosis or port-wine stain and Sturge–Weber syndrome. Table 1 lists neurocutaneous disorders that affect the head and neck, their constituent findings, and genetic causes (1–9).
Eye, Orbit, and Eye Movements Typically, patients with primary eye complaints are seen first by an ophthalmologist. Certainly, patients who present with skull base tumors that affect the orbit, eye movements, or visual pathways should have a rigorous ophthalmologic evaluation. A close working relationship with a neuroophthalmologist is necessary for any skull base team. The following description of the eye examination is provided to highlight the important signs and symptoms to discover and note in patients with skull base tumors.
PHYSICAL EXAMINATION
Eye and Orbit
Due to the compact and complex regional anatomy of the head and neck, the neuro-otologic examination requires practice and experience to refine. As with a general physical examination, a head-to-toe organization is systematic and simple to perform for completeness. In doing so, cranial nerves are checked as one moves from site to site, though a dedicated and rigorous cranial nerve examination requires additional techniques and observation. On initial introduction to the patient, one might identify an obvious facial paralysis or a wet, weak or hoarse voice. Yet, the examining physician should avoid being sidetracked by these overt signs, and instead conduct a systematic review to look for all abnormalities. The reader probably has experienced the patient who has long-standing facial paralysis and whose presenting complaint is totally unrelated to that finding. Additionally, patients frequently present with a
The symmetry of the orbits and eyes should be compared. The globes should be compared to look for proptosis. The status of the lids should be examined not only for lesions but also for their conformity to the globe. Ectropion is very common with facial paralysis, and corrective measures can be performed to minimize its effects. The status of the conjunctiva should be noted for inflammation or irritation, as this too is a frequent sign accompanying facial paralysis. The size and reactivity of the pupils should be documented. This is done both with a flashlight looking for direct and indirect responses and with convergence. Visual fields can be estimated by testing peripheral vision. Lastly, an estimate of vision can be gained by testing each eye separately using a handheld Snellen eye chart that is held about 14 in from the eyes. 95
96 Table 1
Gidley A Listing of Neurocutaneous Disorders That Affect the Head and Neck, Their Congeners, Associated Findings, and Genetic Cause
Diagnosis
Dermatologic manifestation
Ataxia telangiectasia (1,2)
Cutaneous telangiectasia
Neurofibromatosis Type 1 (2,4)
Caf´e au lait spots
Neurofibromatosis Type 2 (5, 6)
PHACE syndrome (7)
Cutaneous and airway hemangiomas
Sturge–Weber syndrome (3,8)
Port wine stain
Tuberous sclerosis (3)
Adenoma sebaceum (facial angiomatosis)
Von-Hippel Lindau (5,9)
Associated findings
Genetic cause
Progressive neurologic deterioration Immunodeficiency High incidence of neoplasms (lymphoid tumors, 10–15%) Multiple neurofibromas Lisch nodules of the iris Optic glioma Bilateral acoustic neuromas Schwannomas of other nerves Meningiomas Ependymomas Gliomas Posterior fossa malformations (Dandy–Walker) Arterial anomalies Coarctation of the aorta and cardiac defects Eye abnormalities Leptomeningeal angioma Choroidal angioma Epilepsy (80%) Learning disabilities Epilepsy (78%) Learning disabilities Giant cell astrocytoma Cardiac rhabdomyosarcoma Lymphangiomatosis Renal angiomyolipoma Cerebellar hemangioblastoma (60%) Retinal angiomas (60%) Renal, pancreatic cysts Renal cell carcinoma (40%) Pheochromocytomas Endolymphatic sac tumors (ELST)
ATM
17q11.2-Neurofibromin
NF2-22q12.2-Merlin
?X-linked dominance ?Developmental disorder
Sporadic ?somatic mutation
TSC1-9q34-hamartin TSC2-16p13.3-tuberin
VHL-3q 11-“VHL protein”
? Indicates possible.
nystagmus (13). In pendular nystagmus, the two phases of nystagmus are of equal length. Jerk nystagmus, which has more relevance in this discussion, has two components: a slow phase followed by a fast phase. Its direction is named for the direction of the fast component. The primary plane or axis of nystagmus is described as horizontal, vertical, rotatory, or direction changing. Classically, horizontal nystagmus is associated with peripheral lesions. The fast component is toward the unaffected ear. Vertical and direction-changing nystagmus are generally signs of central pathology. The time of onset can be described as spontaneous (occurring without provocation), latent (meaning there is some delay in onset, usually after a change in position), gaze-evoked (brought out with certain eye movements), or positional (brought out by certain positions).
Eye Movements Normal eye movements depend on the equal function of the cranial nerves III, IV, and VI and their innervated muscles. A basic test of eye movements includes having the patient follow the examiner’s finger. All of the extraocular muscles are tested in nine different positions (straight ahead, right, left, up, down, as well as diagonally right up right down, left up and left down) (10). The cranial nerves, their muscles, and their actions are noted in Table 2 (11).
Nystagmus Nystagmus is an involuntary, rhythmic movement of the eyes, and the term is derived from the Greek word nystagmos for nodding (12). Generally, nystagmus can be divided into two large categories: jerk nystagmus and pendular Table 2 Eye Movements: The Cranial Nerves, Muscles, Actions, and Deficits Cranial nerve III—oculomotor
Muscle Medial rectus Superior rectus Inferior rectus Inferior oblique
III—oculomotor, parasympathetics VI—superior oblique
Ciliary muscle Superior oblique
VI—abducens
Lateral rectus
Source: Ref. 11.
Action Adduction 1. Elevation 2. Adduction, intorsion 1. Depression 2. Adduction and extorsion 1. Elevation 2. Extorsion and abduction Constriction of the pupil 1. Depression 2. Intortion and abduction Abduction
Deficit Eye is outward, downward, and dilated; upper lid ptosis
Skew deviation, head tilted toward weak side Cannot move eye laterally past midline
Chapter 5: Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base
Spontaneous nystagmus represents an imbalance in the vestibular-ocular reflex (VOR) and can be either central or peripheral in origin (14). Spontaneous nystagmus is best evaluated by examining the eyes through Frenzel lenses. These +10 diopter lenses not only magnify fine movements but also eliminate visual fixation, which could overpower a vestibular nystagmus (14). Spontaneous nystagmus that does not abate with visual fixation probably represents central pathology. Pure vertical, torsional, or linear nystagmus cannot be explained by involvement of a single canal or single labyrinth and implies a central etiology (14). Gaze nystagmus can be elicited by having the patient follow the examiner’s finger performing a “+” type movement, thus evoking either lateral gaze or upward or downward gaze nystagmus. Gaze-evoked nystagmus most often occurs as a side effect of medications or toxins (10). Horizontal gaze-evoked nystagmus usually indicates a lesion in the brainstem or cerebellum; vertical gaze-evoked nystagmus is found in midbrain lesions involving the interstitial nucleus of Cajal (15). Gaze nystagmus has been described as first degree (occurs only with gazing in the direction of the fast component), second degree (occurs in the direction of the fast component and straight ahead), and third degree (occurs in all three directions of gaze). The significance of these distinctions is that first-degree nystagmus is seen with a peripheral lesion and second- and third-degree are seen with central pathology (16). Congenital nystagmus generally beats horizontally at various frequencies and amplitudes and increases with fixation (10). Bruns nystagmus is associated with large posterior fossa tumors. This nystagmus is a coarse, large amplitude horizontal gaze nystagmus toward the tumor side and a fine, high frequency gaze nystagmus away from the tumor side (14). This is thought to be due to compression of the flocculus bilaterally (17). Dynamic vestibular imbalance can be assessed by passive head movement and observation of the patient’s eyes. Dynamic visual acuity is assessed by passively rotating the head at above 2 Hz while the patient reads a Snellen eye chart at the standard distance. A drop in visual acuity of more than one line indicates abnormal gain in the VOR (18). A computerized form of this test has also been developed (19), and might be useful as a clinical tool to separate unilateral from bilateral vestibular hypofunction. Positional nystagmus is provoked by using (Margaret) Dix-(Charles) Hallpike maneuver (20). In this test, the patient is seated on an examining table or bed. The patient is instructed to turn the head to the right side and then to recline backwards as quickly as possible. The examiner watches the patient’s eyes for any nystagmus and asks whether the patient feels dizzy. If nystagmus is present, its length of latency and duration should be noted. Once this nystagmus or dizziness disappears, the patient is asked to return to an upright sitting position. Again the eyes are examined for nystagmus. Patients who have benign paroxysmal positional vertigo will have a latent (usually 2–5 seconds), geotropic (nystagmus beats toward the ground), rotatory nystagmus that reverses (reversibility) when seated upright. The duration and strength of the nystagmus lessens with each subsequent test when performed in repetition (fatigability). Head Thrust Test The head thrust test is a bedside test to assess the VOR (21). The test is performed by asking the patient to focus on a target. The examiner gently grasps the head, and a small am-
97
plitude (5–10 degree), high acceleration (3,000–4,000 degrees/ sec2 ) thrust is applied. The examiner watches the eyes at the end of the head thrust for a corrective saccade. Normal individuals do not use a corrective saccade, the eyes stay fixed on the target. Patients with vestibular hypofunction use a corrective saccade after the head thrust, and the saccade is toward the side of the lesion. This corrective saccade returns the eye to the target and indicates a decreased gain (eye velocity/head velocity) of the VOR (22). The specificity of head thrust test to identify lateral semicircular canal pathology is very high (95–100%), and it correlates 100% with surgical vestibular nerve section (23). In patients with lesser degrees of unilateral hypofunction, the sensitivity is as low as 34% to 39% while the specificity remains high as 95% to 100% (24– 26). Sensitivity of this test can be improved by 30 degrees of cervical flexion (making the horizontal canal horizontal) (22). Head-Shaking Nystagmus Head-shaking nystagmus (27) is used to demonstrate asymmetry in the velocity storage that can occur with either central or peripheral lesions. In this test, the patient’s head is shaken either actively or passively in the horizontal plane for 10 to 15 seconds with the eyes closed. After stopping and opening the eyes, nystagmus will be seen beating away from the side of the lesion (14). In this test, Frenzel lenses are indispensable since the nystagmus is often fine and fleeting. Head-shaking nystagmus can also be performed with electronystagmography, and is discussed in the electronystagmography section.
Ear Pinna and External Auditory Canal The external ear includes the pinna, the membranous external auditory meatus, and the bony ear canal. In one’s zest to examine the eardrum, the pinna might be overlooked. Around 15% to 20% of head and neck skin cancer occur on the external ear, and 55% are along the helix (28–30). The examination should not be limited to just the protruding parts of the pinna, but should also include the postauricular skin as well (Fig. 1). A distinction is made here between the external auditory meatus, which is the cartilaginous outer one-third of the ear canal, and the bony ear canal, which is the
Figure 1 Benign lesion behind left ear.
98
Gidley
Figure 4 Otoendoscopy showing glomus tympanicum tumor. Figure 2 Squamous cell cancer confined to the ear canal. This tumor does not extend past the eardrum into the middle ear and does not extend into the conchal bowl.
bony two-thirds of the external ear canal. The membranous meatus has a thick squamous epithelium, is hair covered, and contains the cerumen glands, while the bony canal is lined by a thin squamous epithelium without any modified sweat glands or hair (Figs. 2 and 3).
Tympanic Membrane and Middle Ear These structures are best evaluated with a handheld otoscope and an otomicroscope. Each modality gives a slightly differ-
ent, but complementary view of the ear canal and tympanic membrane. Each also allows testing of tympanic membrane movement with either a bulb attachment to the otoscope or through a Siegle otoscope viewed through the microscope. Occasionally, patients have a narrowed or oddly shaped ear canal that cannot be adequately evaluated with an oval or round speculum. For this circumstance, using either a pediatric or adult nasal speculum is helpful. Ear endoscopes have been available for the last few decades, but have not found wide use throughout the community. These endoscopes measure 6 cm in length and 2.7 or 4 mm in diameter. Zero- and 30-degree varieties are available. These endoscopes can provide a view of middle ear structures that is unparalleled and not available through a microscope or a handheld otoscope. Additionally, high-quality photodocumentation of ear canal, eardrum, and middle ear pathology can be captured easily with an endoscope (Figs. 4 and 5).
Tuning Fork Examination In a bygone era, tuning fork examination represented the state of the art in audiometric assessment. With the advent of calibrated audiometers, the tuning fork examination has
Figure 3 Squamous cell cancer filling external auditory meatus. This cancer fills much of the conchal bowl but does not extend medially past the bony–cartilaginous junction in the ear canal.
Figure 5 Cholesteatoma of left middle ear as seen with an otoendoscope.
Chapter 5: Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base
lost its centrality in measuring a patient’s hearing. However, it has not lost its importance in giving the attentive clinician valuable information regarding hearing. While the history of otology and audiometry describes several different tuning fork tests, only the Weber and Rinne tests are performed routinely. The Weber test is performed by striking the tuning fork and then placing it on the patient’s forehead, philtrum, or upper incisors and asking the patient where the sound is heard best: right side, left side, or midline (or equally in both ears). The sound is heard better in an ear with a conductive loss or in an ear with better sensorineural hearing when no conductive component is present in the other ear. Interestingly, patients with unilateral, congenital conductive hearing loss do not have lateralization to that side. The Rinne test is performed by striking the tuning fork and then asking the patient to compare for loudness the sound produced with the fork on the mastoid versus in front of the ear canal. A normal result (also called a “positive Rinne”) is that the loudness is greater in front of the ear by air conduction than it is by bone conduction (also noted as AC > BC). In a conductive loss more than about 25 dB with a 512 Hz fork, the sound is louder on the bone than it is through air conduction (noted as a “negative Rinne” or as BC > AC). These tests are easy to perform in the outpatient clinic or at the bedside. They provide a quick assessment of hearing and can be used as a check on audiometric findings (31,32).
Nose and Nasopharynx Anterior rhinoscopy demonstrates the health of the nasal mucosa, the status of the turbinates, and the anatomy of the septum. Boggy, purplish, congested turbinates with thin, clear nasal mucus are often an indicator of allergic or irritant rhinitis. Polyps and other masses are noted. Pathology in the nose should be more closely examined with rigid nasal endoscopes. This allows evaluation of the paranasal sinus meati and permits an adequate measurement of extent of disease. The nasopharynx can be examined indirectly with a heated mirror through the oral cavity; however, only a few patients permit an adequate examination with this approach. The nasopharynx is much better assessed with endoscopes. In this circumstance, the Eustachian tube orifice, posterior and lateral pharyngeal wall, and palate movement can be examined. Palatal myoclonus is best examined with nasopharyngeal endoscopy since the mouth opening required for a peroral examination eliminates this tremor. Fine Eustachian tube endoscopes are developed and are being evaluated. They might help to shed light and treat disorders such as patulous or obstructed Eustachian tube (33).
Oral Cavity and Oropharynx Even a cursory evaluation of the oral mucosa allows examination of the state of the oral mucosa, the teeth, and the tongue. Pathologies in the oral cavity are a frequent cause of referred otalgia. Tongue protrusion and movement from side to side denotes normal hypoglossal nerve function. Fasciculations and atrophy are indicators of abnormal hypoglossal function. The protruding tongue will point to the side of the lesion. The parotid and submandibular ducts are easily assessed by examining their drainage while massaging the respective gland. The oropharynx is bounded by a plane through the hard-soft palate junction superiorly and the circumvallate papillae inferiorly. It contains the palatine tonsils, if still present, lingual tonsils, soft palate, and mucosa of the lateral and posterior pharyngeal walls. The examiner should
99
note the status of this mucosa. Frequently, chronic postnasal drainage will produce cobblestoning of the posterior pharyngeal wall. The movement of the palate should be noted. Unilateral palate weakness, from a glossopharyngeal injury, will allow the uvula to be pulled toward the intact (normal) side. Gag reflex can be elicited by touching the base of the tongue or the lateral pharyngeal walls. A uvula that is bifid or that contains a thin membrana pellucida might be indicator of a submucous cleft palate. This finding warrants digital palpation of the hard palate to assess for occult cleft.
Larynx Laryngeal assessment begins with the interview, noting the patient’s voice quality. Wet, weak or breathy voices are signs of vocal fold weakness. Speech fluency, on the other hand, is directed by higher, cortical structures. Expressive aphasia from a lesion in Broca area is an example of an abnormal fluency. Dysarthria or difficult, poorly articulated speech might be due to abnormal hypoglossal or facial nerve function. A basic examination of laryngeal function includes mirror examination; however, only an exceptional patient permits an unhurried examination with this technique. Flexible fiberoptic endoscopy under topical nasal anesthetic is the preferred initial method to assess vocal fold function. Patients who have or will have vocal fold paralysis should be evaluated by a speech pathologist. Videostroboscopy provides an excellent assessment of vocal fold anatomy and function, and can discern various degrees of weakness better than flexible endoscopy. Additionally, videographic and photographic documentation of vocal fold function and appearance is much better with videostroboscopy than through a flexible fiberoptic scope. Laryngeal function is controlled by the vagus nerve. The vagus nerve is a mixed nerve carrying motor, sensory, and parasympathetic impulses. Its first branch in the neck is the superior laryngeal nerve, which has two branches: an internal laryngeal branch that carries sensation from the mucosa above the true vocal folds and an external branch that is motor to the cricopharyngeal muscles. This muscle tilts the thyroid cartilage on the cricoid cartilage and produces the tightening of the vocal fold needed to make high-pitch phonation. The remainder of the laryngeal muscles and the sensation of the vocal folds and mucosa of the tracheobronchial tree are innervated by the vagus and its inferior or recurrent laryngeal nerve. This nerve branch loops under the arch of the aorta on the left side of the neck and the subclavian on the right side. Nonrecurrent nerves (nerves that do not descend into the mediastinum before going to the larynx) have been well described. This situation occurs more commonly on the right side, and is constantly on the minds of thyroid surgeons. The recurrent laryngeal nerve innervates both the adductors and abductors of the vocal folds. For this reason, recovery from laryngeal neurotmesis might be limited due to synkinesis of laryngeal innervation. Injury or loss of vagal (and for that matter glossopharyngeal) function at the skull base can produce severe dysphagia and aspiration since both the motor and sensory functions lost, resulting in loss of the protective mechanisms of the upper aerodigestive tract.
Neck, Parotid, and Thyroid The neck and, by extension, the parotid glands should be examined by palpation for any masses or lymphadenopathy. Cervical adenopathy should be reported based on its
100
Gidley
location and size. For malignant disease, location and size of lymphadenopathy is important for tumor staging. The parotid glands should be palpated for any masses, especially in patients with facial paralysis. Loss of a single branch of the facial nerve is due to malignant involvement until proven otherwise. Complete facial paralysis can occur from parotid tumors at the stylomastoid foramen. Deep lobe tumors can also cause facial paralysis and can escape palpation; only imaging studies can show these tumors. The thyroid gland sits on top of the trachea, just below the level of the cricoid cartilage and above the sternal notch. Asking the patient to swallow while one palpates this part of the neck moves the gland under the examiner’s fingers. A normal gland is usually not able to be palpated. A solitary tumor should be further investigated with ultrasound and fine needle aspiration. Diffuse swelling of the gland might indicate goiter. Vocal fold paralysis associated with a thyroid nodule is due to malignant disease until proven otherwise. Lastly, the spinal accessory nerve, CN XI, function is tested. This nerve innervates the sternocleidomastoid (SCM) and trapezius muscles. To test the SCM muscle, the patient is asked to turn the head slightly to the right while the examiner applies an opposite force against the right jaw and face and palpates the strength in the left SCM. The head is turned slightly to the left to test the right SCM muscle strength. The patient should be asked to shrug shoulders while the examiner applies resistance to measure trapezius strength.
Cranial Nerves “Any symptom suggestive of cranial neuropathy must alert the clinician to the possibility of a skull base or intracranial space occupying lesion” (34). For the most part, the cranial nerve examination is performed as one progresses through the head and neck physical examination. Eye, palate, tongue, vocal fold, and shoulder/head movement have already been discussed in those sections related to each organ system. This space does not permit a discussion of the complex anatomy of the cranial nerves, their brainstem, and skull base relations. Some of this detail has been summarized in Table 3 (35–37). Here, the nerves that are not readily examined in a routine examination are described to be performed as part of a neuro-otologic examination.
Olfactory Nerve Although taste and smell provide great sensory stimulation, they are perhaps the least tested sensory functions (38). The sense of smell occurs when odorants come into contact with olfactory receptors in olfactory receptor neurons (39). The olfactory receptor neurons are bipolar cells that penetrate the cribriform plate to synapse with glomeruli cells in the olfactory bulb. These glomeruli cells then synapse with mitral cells that carry the signal into the piriform cortex. From the piriform, connections with the hippocampus, amygdala, and the orbitofrontal cortex combine to give the associated sensory, memory, or hedonic reactions (39). Tests on olfaction are becoming more important as their significance in predicting Parkinson disease and Alzheimer disease is recognized (39,40). Anosmia or the inability to smell can be evaluated based on conductive or sensorineural causes (41). Olfactory epithelium is located high inside the nose on the upper middle turbinate and roof of the nose. Odorants must be able to pass through a patent nasal airway, and the patient must be able to generate enough nasal airflow (sniff) to bring odorants into contact with the olfactory epithelium. The patient
with severe obstructive nasal polyposis represents a form of conductive anosmia since the polyps prevent any airflow through the nose. Furthermore, a laryngectomee is unable to move any air through the nasal passages since all of the airflow is through the tracheostoma. Sensorineural anosmia can be due to either damaged olfactory mucosa as might occur following a viral upper respiratory tract infection or shearing of the olfactory nerve as a result of head trauma. Tumors in the sinonasal tract and esthesioblastomas can produce anosmia through obstruction and/or destruction along the olfactory pathway. Olfactory loss from intracerebral tumors (glioma, olfactory meningioma, esthesioneuroblastoma, and adenocarcinoma) has been described (39). For this reason, patients with anosmia and a normal nasal endoscopic examination should undergo MRI imaging (39). Olfaction can be measured in several different ways. A simple test of smell might include using household products such as coffee, cinnamon, mint, or water and asking the patient to respond to the question of “Do you smell this item?” with a “yes” or “no” answer when the item is waved under the nose and closed eyes (13,42). Eighty percent of the normal population can identify the odor of coffee (34). A patient who professes not to smell anything can be further tested with ammonium in a similar fashion. Ammonium triggers response from V2 since it is a mucosal irritant. All normal patients sense the nasal irritation of ammonia (34). A “no” response in this circumstance might indicate that the patient is being less than honest in his responses. A better form of testing uses the University of Pennsylvania Smell Identification Test (43). This test uses 40 scratch and sniff odors that patients try to identify by matching to a four-option response panel. Patients are required to answer all questions and to guess at an answer even if they do not smell anything. The number of correct answers is matched in a normogram divided by gender and age range to determine the level of olfaction: normal, hyposmia, anosmia, malingering. Since there is always a one in four chance of getting any question correct through guessing, a patient who has missed all 40 questions has probably intentionally avoided correct answers (40). This test is performed with both nostrils open and thus cannot give side specific values. Nonetheless, this method of testing has been widely verified and is used for its accuracy in providing a qualitative measurement of smell (41).
Trigeminal Nerve The trigeminal nerve is perhaps the most complex nerve of the head and neck. It has origins from four different brainstem nuclei; it carries both sensory and motor functions; its three main trunks pass through three different foramina in the skull base; and each trunk has an associated parasympathetic ganglion. The trigeminal nerve has general somatic and general visceral afferent function. The somatic afferents are for cutaneous sensation of the skin of the head and neck. The first division, V1 , supplies sensation to the cornea and conjunctiva, the upper eyelid, the eyebrow, and the scalp as far posterior as the vertex. The second division, V2 , supplies sensation to the skin of the nose, cheek, upper lip, to the maxillary teeth, and to the mucosa of the nose, and the roof of the mouth. The third division, V3 , provides sensation to the skin of the lower lip, chin, and lower one-third of the face; to the mandibular teeth; and to the mucosa of the cheeks, lower lip, and floor of mouth. The general somatic afferents of CN V are also carried on V3 to the anterior two-thirds of the tongue. Light touch,
None (ganglion cell layer in retina, optic tracts project to lateral geniculate body) Oculomotor nucleus (lateral somatic cell column) Oculomotor nucleus (caudal central group)
II—optic
Principal sensory, geniculate ganglion Descending, spinal tract and nucleus of V, geniculate ganglion Mesencephalic nucleus of V (no peripheral ganglion) Abducens nucleus Facial motor nucleus Intermediate nerve, geniculate ganglion, spinal trigeminal tract Intermediate nerve, geniculate ganglion, nucleus solitarius (“gustatory nucleus”) Superior salivatory nucleus, intermediate nerve
Cochlear nuclei (dorsal and ventral)
V—trigeminal, sensory (GSA) V—trigeminal, sensory (GSA)
VII—facial, sensation (GSA)
VII—facial, parasympathetics (GVE)
VIII—cochlear nerve
VII—facial, sensation (SVA)
VI—abducens (GSE) VII—facial, motor (SVE)
V—trigeminal, sensory (GSA)
Motor nucleus of V
Superior cervical ganglion of sympathetic chain Trochlear nucleus
V—trigeminal, motor (SVE)
IV—trochlear, motor (GSE)
III—oculomotor, parasympathetics (GVE) III—oculomotor, sympathetics
III—oculomotor (SVE) Edinger–Westphal nucleus
None (olfactory glomerulus in olfactory bulb)
I—olfactory
III—oculomotor
Brainstem nucleus
Cranial nerve
Table 3 Cranial Nerves, Their Nuclei, Brainstem Location, and Function
Pons
Pons
Pons
Pons
Pons Pons
Midbrain
Pontine tegmentum Pons to C2-C4
Midbrain (level of inferior colliculus) Pontine tegmentum
Midbrain (level of superior colliculus) Midbrain (level of superior colliculus) Midbrain (level of superior colliculus)
None (olfactory tract to anterior perforated substance and pyriform lobe) None (diencephalon)
Brainstem location
Sensation of taste for anterior 2/3 of the tongue Lacrimal gland and nasal mucosa via GSPN and PtPG (V2 ) Submandibular gland via chorda tympani (and lingual nerve, V3 ) Hearing
Motor to lateral rectus mus Muscles of facial function, buccinators, platysma, stapedius mm Sensation of ear canal
Proprioception from the TMJ
Muscles of mastication, accessory muscles of mastication (tensor tympani, tensor veli palatini) Light touch and pressure Pain and temperature
Constrictor of pupil Mueller muscle in upper eyelid Superior oblique
Dilator of the pupil via NCG of V1
Levator palpebrae
Extraocular muscle movement
Vision
Olfaction
Function
Deafness
?dry eyes, nose, mouth
(Continued)
Loss of taste, anterior 2/3 of the tongue
Diplopia, eye does not abduct Facial paralysis, loss of stapedial reflex (hyperacusis) Numbness of ear canal (Hitzelberger sign)
Loss of jaw jerk reflex
Loss of sensation in distribution Loss of sensation in distribution
Vertical diplopia (trouble walking down stairs) Weakness of muscles of mastication
Horner syndrome
“Blown pupil”
Diplopia, outward and downward deviation of eye Ptosis
Blindness
Anosmia
Deficit
Chapter 5: Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base
101
Hypoglossal nucleus
Spino-medullary junction
Spino-medullary junction Medulla and C1–C3
Spino-medullary junction
Spino-medullary junction
Spino-medullary junction
Spino-medullary junction
Motor to the tongue
Parasympathetic to the GI tract Motor to SCM and trapezius
Hypopharynx, larynx, and tracheobronchial tree Taste, tip of epiglottis
Parotid gland via otic ganglion (lesser petrosal nerve and auriculotemporal br of V3 ) Vocal fold movement, pharyngeal constrictors, cricopharyngeus Postauricular skin and ear canal
Dysarthria
Weakness of shoulder
Lack of laryngeal taste
Dysphagia, aspiration
Numbness
Hoarseness, dysphagia, aspiration
Lack of pharyngeal taste, loss of carotid sinus reflex ?xerostomia
Dysphagia
Loss of gag reflex
Numbness
Imbalance
Deficit
Abbreviations: GSA, general somatic afferent; GSE general somatic efferent; GVA, general visceral afferent; GSPN, greater superficial petrosal nerve; NCG, nasociliary ganglion; PtPG, pterygopalatine ganglion; SCM, sternocleidomastoid; SVA, special visceral afferent; SVE, somatic visceral efferent. Source: Refs. 35–37.
X—vagus, parasympathetics (GVE) XI—spinal accessory, cranial root (SVE) XI—spinal accessory, spinal root (GSE) XII—hypoglossal (GSE)
X—vagus, special sensation (SVA)
Rootlets from C1–C3
Superior ganglion, in neck, spinal trigeminal tract Inferior (or nodal) ganglion in neck, nucleus solitarius Inferior (or nodal) ganglion in neck, nucleus solitarius Dorsal motor nucleus Nucleus ambiguus
X—vagus, sensation (GSA)
X—vagus, sensation (GVA)
Nucleus ambiguus
Medulla
Inferior salivatory nucleus
Medulla
Nucleus ambiguus Medulla
Sensory to base of tongue and upper pharynx Motor to stylopharyngeus and superior constrictor mm Taste, base of tongue, baroreceptor
Medulla
Nucleus solitarius (“gustatory nucleus”)
Sensation to postauricular skin
Balance
Function
Medulla
Pons
Brainstem location
Vestibular nuclei (superior, medial, lateral, inferior) Superior ganglion, in neck, spinal trigeminal nucleus Inferior salivatory
X—vagus, motor (SVE)
IX—glossopharyngeal, special sensory (SVA) IX—glossopharyngeal, parasympathetic (GVE)
IX—glossopharyngeal, sensory (GSA) IX—glossopharyngeal, sensory (GVA) IX—glossopharyngeal, motor (SVE)
VIII—vestibular nerves
Brainstem nucleus
Cranial Nerves, Their Nuclei, Brainstem Location, and Function (Continued)
Cranial nerve
Table 3
102 Gidley
Chapter 5: Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base
pinprick, and temperature can be tested for each division of the trigeminal nerve. The motor function of the trigeminal nerve is carried on V3 to the masticatory muscles (masseter, temporalis, medial, and lateral pterygoid) and to the accessory muscles of mastication (mylohyoid, anterior belly of the digastric, tensor tympani, and the tensor veli palatini). Motor strength is difficult to assess clinically, but the tone of the masseter and temporalis can be palpated while the patient grinds teeth. Corneal Reflex The description of the corneal reflex rightfully belongs between the discussion of trigeminal and facial nerve function since it involves both nerves. The corneal reflex arc is composed of afferents from corneal epithelium through the V1 into the brainstem, where synapses through one or two interneurons and connects to the facial nuclei to produce muscular contraction of the orbicularis oculi muscles. Normally, unilateral corneal irritation produces bilateral orbicularis oculi contraction. The test is performed by asking the patient to look slightly nasally while a wisp of cotton is placed on the temporal portion of the cornea. Care is taken to prevent the patient from seeing the cotton approach the eye. In the case of a trigeminal nerve lesion, no muscular contraction will be elicited in either eye. In the case of a facial nerve lesion, the muscular contraction will be absent on the side of the lesion but will still be present on the normal side.
Facial Nerve As the motor supply to the face, the facial nerve controls several different, wide ranging functions. Under control of the facial nerve, the buccinator muscle aids in mastication by helping to keep the food bolus on the occlusal surface of the molars and not in the gingivobuccal sulcus. The perioral muscles aid in articulation of speech (for labial and plosive sounds). The orbicularis oculi protects the eye by closing the lids and by moving a lubricating coating of tears over the cornea. Lastly, and perhaps most importantly, it allows nonverbal communication of emotion through facial muscle contractions. Like the trigeminal nerve, the facial nerve has an intricate anatomy including the longest bony course of any cranial nerve, conveyance of motor, general sensory, and special sensory function, and is the primary pathway for two parasympathetic ganglia (sphenopalatine and submandibular). Since the facial nerve has many different branches along its course, each with a testable function, a topographic method of testing had previously been used to determine the site of lesion. In this way, testing lacrimation (Schirmer test), taste (electrogustometry), salivation, stapedial reflexes, and facial muscle function can provide the examiner with the location of the lesion. However, this topographic testing has largely been supplanted by modern imaging techniques and thus is not widely used. Acoustic reflex testing is performed and will be discussed in the section on audiometric testing. Facial muscle function is tested by asking the following to the patient: “raise your eyebrows,” “close your eyes tightly” even against resistance, “wrinkle your nose,” “puff out your cheeks,” “pucker your lips,” “show me your teeth” [Fig. 6(A)–6(D)]. In case of complete paralysis, the examiner presses his thumbs on the midline of the patient’s face to prevent the unopposed normal side from distorting the examination of the paralyzed side. Facial function should be reported for each area of the face, since just one branch might be paralyzed. Single branch
103
facial paralysis strongly suggests tumor (44). Any individual with progressive facial weakness should be considered to have a tumor of the facial nerve until proven otherwise (44). The American Academy of Otolaryngology has approved the House–Brackmann scale as a measure of facial function for reporting results in its publications (Table 4) (45). The constellation of cranial nerve abnormalities can be grouped along recognizable patterns. These patterns, often carrying eponyms, can help in deducing the underlying pathologic process. Several such syndromes are listed in Table 5 (41,46–52).
Gait and Balance Testing A general neurologic examination of upper and lower body strength and sensation and cerebellar function should be performed looking for light touch, vibratory sensation, fine and rapid motor skills, and finger-to-nose testing (13).
Gait Patients are asked to walk for 15 ft and to return in order to assess the gait. The observer should note the posture, the position of the head, the motion in the arms, and the gait. Foot drop or poor posture might be indicators of imbalance. One should note how the patient turns himself around: does he use a quick method without needing to stop or does he make several small steps to turn himself around? The latter might be an indication of a vestibular pathology. Next, the patient is asked to walk heel to toe (tandem gait) for 15 ft. The examiner should accompany the patient to be certain that the patient does not fall. An abnormal tandem gait is loss of balance more than three times in 15 ft.
Stance Moritz Heinrich Romberg first published his description of tabes dorsalis in 1846 (53). Romberg test is performed with eyes closed, feet together, arms folded across the chest, and head extended (54). A sharpened Romberg test is performed with feet tandem (heel to toe), eyes closed, arms folded across the chest, and head extended. A simplified Romberg is performed with feet together, arms at the side, head in a neutral position, and eyes closed. The examiner looks for increased body sway and protects the patient from falling. This can be done by asking the patient to stand between two chairs, with his back about 50 cm from the wall. Normal individuals can maintain a Romberg posture for 30 seconds with eyes closed. Romberg posture relies on normal proprioception (dorsal columns), and an abnormal Romberg often indicates disease outside of the labyrinths (54). Patients are then asked to stand on one leg with eyes open and then eyes closed. Normal patients can easily stand on one leg for 15 seconds. Inability to stand on one leg indicates imbalance and highlights the need for normal muscle and joint strength and normal proprioception to maintain balance.
Fukuda Stepping Test The Fukuda (55), also called Unterberger (56), stepping test is performed with eyes closed, arms outstretched while the patient marches in place. The examiner watches the patient’s movement during 50 marched steps. Forward movement (>1 m) or turning movement (>45 degrees, usually toward the side of the lesion) are significant findings in the Fukuda test (57). This is a test of vestibulospinal and proprioceptive contributions for balance control (58). Some reports have shown good sensitivity of this test (57), while others have discounted its usefulness (56,59).
104
Figure 6 paralysis.
Gidley
(A)
(B)
(C)
(D)
Facial function (A) at rest, (B) during raised eyebrows, (C) during closed eyes, (D) while showing teeth in a patient with complete right facial
Clinical Test of Sensory Integration and Balance This series of tests uses simple objects to test static and dynamic equilibrium (54,60). Romberg postures with eyes open and closed are used to mimic the test situations 4 and 5 found in computerized dynamic posturography (54). Thick upholstery foam, a rigid cover, and lampshade are used to create the “foam and dome” test modalities. These simple clinical tasks of static and dynamic equilibrium can reliably distinguish vestibular disorder patients from normal subjects (61).
AUDIOMETRIC, VESTIBULAR, AND ELECTROMYOGRAPHIC STUDIES These tests provide valuable insight into disease process and help to measure function. The following text is not meant to describe how to perform each test. Where indicated, a rudimentary description will be given to familiarize the reader with how the test is performed, but more importantly, how to interpret in the findings and
Chapter 5: Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base
105
Table 4 House–Brackmann Facial Nerve Grading System Grade
Description
I II
Normal Slight dysfunction
III
Moderate dysfunction
IV
Moderately severe dysfunction
V
Severe dysfunction
VI
Total paralysis
Characteristics Normal facial function in all areas Gross: slight weakness noticeable on close inspection; may have very slight synkinesis At rest: normal symmetry and tone Motion Forehead: moderate to good function Eye: complete closure with minimum effort Mouth: slight asymmetry Gross: obvious but not disfiguring difference between two sides; noticeable but not severe synkinesis, contracture, and/or hemifacial spasm At rest: normal symmetry and tone Motion Forehead: slight to moderate function Eye: complete closure with effort Mouth: slightly weak with maximum effort Gross: obvious weakness and/or disfiguring asymmetry At rest: normal symmetry and tone Motion Forehead: none Eye: incomplete closure Mouth: asymmetric with maximum effort Gross: only barely perceptible motion At rest: asymmetry Motion Forehead: none Eye: incomplete closure Mouth: slight movement No movement
Source: Ref. 45.
Table 5 Cranial Nerve Syndromes Syndrome Foster Kennedy syndrome (41)
Cranial nerves involved
Orbital apex (46)
I (ipsilateral hyposmia or anosmia) II (ipsilateral optic atrophy and central papilledema) III, IV, VI, V1
Cavernous sinus (46,49)
III, IV, VI, V1 , V2 + sympathetics
Superior orbital fissure (Rochon–Duvigneaud or Foix) syndrome (46, 47) Retrosphenoidal space (Jacod) (47) Petrous apex (Gradenigo syndrome) (47)
III, IV, V1 , VI
Miller–Fisher syndrome Internal auditory canal (47)
VI, VII VII, VII
Cerebellopontine angle (47,48)
V, VII, VIII, IX, X, XI
Jugular foramen (Vernet) (47,48,50) Schmitt (48) or Collet–Sicard syndrome (50–52) Retropharyngeal syndrome of Villaret (51)
IX, X, XI IX, X, XI, XII without Horner syndrome IX, X, XI, XII with Horner syndrome
II, III, IV, V, VI V, VI
Most likely cause(s) Tumor of olfactory grove and sphenoidal ridge
Inflammatory (sarcoid, SLE, Wegener’s, Graves) Infectious (fungal, bacterial, viral) Tumors (nasopharyngeal, adenoid cystic, squamous cell, lymphoma) Iatrogenic (sinonasal or orbitofacial surgery) Vascular (carotid aneurysm, carotid cavernous fistula) Mucocele Idiopathic (Tolosa–Hunt) Thrombosis Sellar tumors Trauma Tumors of the middle fossa Chronic otitis Cholesterol granuloma Chondrosarcoma Rhombencephalitis (Herpes, Guillain–Barre) Acoustic neuroma Meningioma Epidermoid tumor Meningioma Acoustic neuroma Epidermoid tumor Neoplasm Carotid aneurysm
106
Gidley
place the results into the framework of the entire clinical picture. Electrodiagnostic testing is a wide field of medical practice. Audiologists, neurologists, neuromuscular specialists, and physiatrists perform one or more of these tests as part of their regular practice and are the recognized experts in performing these tests. Liberal use of these consultants is necessary in the evaluation of patients with skull base tumors. The allotted space does not permit for this chapter to be an exhaustive resource; and therefore, certain tests are eliminated though they might provide insight into disease processes (e.g., electrogustometry). The bulk of medical literature relating audiometric testing and skull base tumors concerns the diagnosis of acoustic neuroma (AN). ANs account for 5% to 10% of intracranial tumors and 80% to 90% of posterior fossa tumors (62). Since Cushing first recognized hearing loss as the presenting symptom of ANs (63), scientists and physicians have tried to develop better audiometric test to identify the presence of ANs. The history of neuro-otologic diagnosis for ANs has included reflex decay, alternate binaural loudness, and B´ek´esy audiometry (64, 65), but these modalities have been replaced by auditory brainstem response (ABR) and MRI. The following topics will primarily relate the findings of audiometry and vestibular testing to ANs. The findings of other tumors, such as meningiomas, epidermoids, etc. will be presented where significant differences are found.
Basic Audiometry The audiogram is the most fundamental element of otologic and neuro-otologic evaluation beyond the history and physical examination. A basic audiogram consists of pure tone audiometry, speech audiometry, immittance testing (e.g., the tympanogram), and acoustic reflex testing. Only a brief description of each test is permitted in this chapter;
the interested reader is directed to other reference works for an in depth discussion of the finer points of these tests (66–69). Pure tone audiometry is the measurement of the lowest threshold at which a tone is heard. A calibrated audiometer delivers sound at a specific frequency (pitch) and specific intensity (loudness). The test is performed by an audiologist in a sound proof booth using either circumaural headphones or ear inserts for air conduction levels and using a bone vibrator for bone conduction levels. Masking sound is given to the nontest ear via an ear canal insert and is the physiologic equivalent of covering one eye during a vision test. The results of pure tone tests are placed on the audiogram using conventional symbols to designate the ear, the modality (air or bone conduction), and the use of masking. The intensity levels of air conduction at 500, 1000, and 2000 Hz (and occasionally 3000 or 4000 Hz) are averaged producing a pure tone average (PTA). Hearing loss can be grouped into three different categories: conductive, sensorineural, or mixed. Conductive hearing loss indicates that the sound conducting mechanism of the ear is impaired. This condition can occur from any process that blocks the ear canal or impairs the vibration of the tympanic membrane or ossicles. Cholesteatoma, cerumen impaction, otitis media, tympanic membrane perforations, ear canal cancers (Fig. 5), and otosclerosis are common causes of conductive hearing loss (Fig. 7). Sensorineural hearing loss indicates that the defect in hearing lies either within the cochlea or the auditory nervous pathway. Common examples of sensorineural hearing loss occur in presbycusis, noise-induced hearing loss, ototoxicity, and ANs (Figs. 8 and 9) Mixed hearing loss means that both conductive and sensorineural hearing loss types are present in an ear.
Figure 7 Audiogram showing left conductive hearing loss due to cholesteatoma. The open circles denote the hearing level tested by ear canal inserts and denote the air conductive hearing level. The solid triangles denote the hearing level tested by a bone vibrator behind the left ear and masking noise in the right ear and mark the left ear cochlear hearing level.
Chapter 5: Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base
107
Figure 8 Audiogram from patient with left intracanalicular acoustic neuroma. The shaded area shows the normal hearing range. There is left sensorineural hearing loss above 1 kHz. The right ear hearing is essentially normal. The word recognition score phonetically balanced maximum (PBM) is 92% for each ear. The tympanograms are normal Type A bilaterally. The auditory reflexes are present bilaterally.
A threshold for perceiving words can be achieved by using spondaic words. The lowest level that 50% of words are perceived is called the speech reception threshold (SRT). The PTA and SRT should agree within 5 to 10 dB of each other. Speech audiometry measures word understanding. Phonetically balanced words are presented via air conduction at a presentation level 40 dB over the SRT or PTA (also called 40 dB sensation level or 40 dB normal hearing level). The percent of words that are understood is recorded as the speech discrimination score (SDS). Generally speaking, word understanding improves as the intensity is increased for cochlear or sensory hearing loss. However, retrocochlear hearing loss might demonstrate a worsening of word understanding with increased intensity, and this is called phonetically balanced rollover. A compilation of differences between sensory or cochlear hearing loss and retrocochlear or neural hearing loss is presented in Table 6 (70–74). Prior to ABR and MRI, site of lesion testing was of paramount importance in discerning patients who might have an AN; however, since up to 20% of AN patients might demonstrate a cochlear rather than a retrocochlear pattern of hearing loss, retrocochlear audiometric testing is of little benefit (75). The American Academy of Otolaryngology Committee on Hearing and Equilibrium (76) composed a classification system for reporting hearing results in AN surgery. This classification system uses PTA and SDS to stratify patients into four different classes of hearing level (Table 7). Alternatively, some authors report hearing results as “unchanged,” “serviceable,” “measurable,” or “not measurable.” In this context, serviceable is PTA ≤ 50 dB and SDS ≥ 50%, unchanged is hearing within 15 dB PTA and 15% SDS of preoperative levels, measurable is any other hearing, and not measurable is self-explanatory (77).
Immittance testing uses an impedance bridge to measure changes in tympanic membrane compliance. Compliance of the tympanic membrane is affected by perforations, middle ear fluid or tumor, and the reflex contraction of middle ear muscles. Several important findings can be discovered with this type of testing. Immittance testing gives a clue regarding the status of the tympanic membrane (intact, perforated, or floppy) and the status of the middle ear (aerated or fluid filled). These findings are denoted on a tympanogram; however, for the discussion of skull base tumors, tympanograms have little significance unless the tumor or spinal fluid invades the middle ear and produces a flat tympanogram. Acoustic reflex testing, on the other hand, has more significance for neuro-otologic diagnosis of skull base tumors. Using the impedance bridge, compliance of the tympanic membrane can be measured in response to a tone burst given either ipsi- or contralaterally. In response to loud sound (85– 110 dB), a reflex contraction of the stapedial muscle will occur bilaterally. A normal reflex requires an intact tympanic membrane, an air-filled middle ear, normal movement of the ossicles, no worse than 35 dB hearing loss, and an intact facial nerve (stapedial muscle). A defect anywhere along this pathway can produce an absent or reduced acoustic reflex. The sensitivity of acoustic reflex testing for ANs has been quoted anywhere between 21% and 90% (75,78–84). Acoustic reflex decay is defined as a 50% loss of middle ear contractility in response to a tone administered 10 dB above threshold. The sensitivity of reflex decay has been reported from 36% to 100% for ANs (75,78,79,81–85).
Acoustic Neuroma Hearing loss is found in up to 95% of patients with an AN (75). By that same token, normal hearing is reported in 3% to 12% of AN patients (75,81,86–88). The current level of
108
Gidley Table 7 AAO–HNS Hearing Classification Class
PTA
SDS
≤30 dB >30 dB, ≤50 dB >50 dB Any level
A B C D
≥70% ≥50% ≥50% ≤50%
Source: Ref. 76.
(A)
(B)
Figure 9 (A) Contrast-enhanced axial MRI and (B) contrast-enhanced coronal MRI of left intracanalicular acoustic neuroma from patient whose audiogram is depicted.
clinical detection of AN is approximately 1 in 100,000 person per year (89), though vastly higher numbers of tumors must be present and escaping detection given the 1% observed rate of acoustic tumors found at autopsy (90). Indeed, a significant number of ANs are found serendipitously on MRI performed for unrelated complaints (91). In general, the degree of hearing loss is significantly linked with tumor size, so that up to 33% of intracanalicular tumors are associated with normal
hearing (83). However, there are many reports of individual large tumors (>2 cm) associated with normal hearing and small tumors (<1 cm) associated with anacusis (92). Schuknecht has calculated that 75% of nerve fibers need to be destroyed before pure tone hearing is affected, given an intact organ of Corti (93). The distribution of high-frequency nerve fibers on the periphery and low-frequency fibers centrally in the acoustic nerve accounts for high frequency hearing loss that is found in early acoustic tumors development. Hearing deteriorates by as much as 2.4 dB per year while ANs are observed (94). Speech discrimination also significantly deteriorates over time in observed ANs (94). In the series of tumors described by Selesnick and Jackler (88), high-frequency asymmetry at 4 kHz was a more sensitive indicator of an AN than difference in either SRT or SDS. While the classic presentation of an AN is a unilateral progressive sensorineural hearing loss with poor speech discrimination (95), experts have not agreed on what exactly constitutes a significant asymmetry (75,96,97). As a rule of thumb, a significant asymmetry in hearing is described as an interaural SRT difference greater than 15 dB, an interaural SDS difference greater than 12% to 20% or an interaural 4 kHz difference greater than 15 dB (75). Obholzer et al. (98) sought to define appropriate audiometric criteria for referral for MRI. They reviewed 392 MRIs performed in 1 year. Of these, 36 patients had ANs; these 36 ANs and 92 randomly selected “normals” were included for the analysis. Audiometric data and clinical histories were evaluated to look for findings that might be indicative of AN. They used the published protocols of seven different studies to analyze audiometric data. Their study supports the use of interaural asymmetry at two neighboring frequencies of >15 dB if the mean threshold in the better ear was ≤30 dB (unilateral hearing loss) and an interaural difference of 20 dB if the mean threshold is greater than 30 dB in the better ear (bilateral asymmetric hearing loss). These criteria had a 97% sensitivity and 49% specificity for AN. The most sensitive individual frequency asymmetry was a 15 dB at 2 kHz, with a sensitivity of 91% and a specificity of 60%. The most sensitive criterion was a difference of 15 dB at any frequency (sensitivity 100% and specificity 29%). Several authors have examined hearing levels as a predictor of hearing preservation in AN surgery. In a multivariate logistic analysis of preoperative hearing variables
Table 6 A Compilation of Differences Between Sensory (Cochlear) vs. Neural (Retrocochlear) Hearing Loss
Site of pathology
Pure tone audiometry (72)
Sensory or cochlear
Hair cell
Decreased
Neural or retrocochlear
Auditory nerve
Relatively unaffected
Speech discrimination (72) Preserved, until widespread cochlear damage Decreased
Secondary findings (71,73) Recruitment
Auditory fatigue or tone decay
Speech discrimination (72) Preserved, until widespread cochlear damage Decreased
ABR (70)
OAE (74)
Intact
Absent
Abnormal or absent
Intact
Chapter 5: Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base
predictive of hearing preservation, Robinette et al. (99) examined the audiometric test results in 104 AN patients. Only word recognition score (WR40) was found to be a significant determinant after accounting for small tumor size (≤2.0 cm). Additionally, they found that patients who had hearing preserved had a higher rate of normal acoustic reflexes than in those patients who did not have hearing preserved (99).
Other Tumors In 1997, Baguley et al. (100) published their series of cerebellopontine angle (CPA) meningiomas and performed a review of the literature. In their series, 80% (20/25 patients) had abnormal pure tone testing and 50% (10/20) had abnormal SDS (i.e., <90%). Interestingly, the five patients who had normal audiometry had large tumors (two in 2.5–3.4-cm range, and three >4.5 cm); equally, 9/10 patients with normal SDS were found to have large tumors (2.5 cm and larger). When combined with the other series reviewed, 37/61 (61%) patients had abnormal PTA and 22/42 (52%) had abnormal SDS (100). Doyle and De La Cruz (101) reported audiometric results in 13 patients with CPA epidermoid tumors. Four patients had PTA greater than 30 dB; SDS was reduced out of proportion to pure tone hearing. Quaranta et al. (102) described the audiometric features in a report of 11 CPA epidermoid tumors. Their series consisted of tumors that measured 3.5 to 7 cm in maximum diameter. They found symmetric hearing in six patients. Another four had asymmetric hearing loss, worse on the tumor side. In a report on 10 epidermoid tumors, Kaylie et al. (103) found normal hearing in three, while the remainder had varying levels of hearing loss from mild to anacusis.
Auditory Brainstem Responses In the history of methods to diagnose ANs, ABR represented a giant advance and promised a much less invasive test. Prior
+ A1
I
to its development, audiologists developed many different tests to stress the auditory nerve to determine its function. But as the preceding paragraphs indicate, many of these tests lacked the specificity or sensitivity necessary to identify tumors. However, once a patient was identified as possibly having an AN, the next step would have required either an air-contrast or a pantopaque posterior fossa myelogram or polytomograms of the internal auditory canal. These tests were not only invasive, but they were extremely painful and potentially morbid. They were not recommended lightly. In 1971, when ABR was introduced, its significance for AN screening had not yet been realized (104); however, before the end of the decade, the sensitivity of ABR to identify ANs was well established (105). Unfortunately for ABR, its heyday was relatively short lived. MRI with gadolinium contrast enhancement was introduced in 1988 (96), and MRI has been nearly 100% sensitive for ANs as small as 4 mm (106,107). ABR is performed with a ground electrode on the vertex and another electrode on the earlobe or mastoid of the stimulated ear. Clicks or tone burst are given at 20/sec or faster rates. Click stimulus estimates hearing in the range of 1000 to 4000 Hz. Intensity can be varied but is generally given at 70 dB; however, with lower intensities the amplitude response decreases and latency increases. Bandpass filters are set from 30 or 100 to 3000 Hz and are used to encompass the spectrum of response while reducing undesirable activity. ABR is influenced by age, gender, and body temperature, but it is not affected much by state of arousal or sedative medications (108). By convention, ABR waveforms are number I through V. Each numeral indicates a positive waveform. These waveforms have been correlated with structures within the auditory pathway: I for the distal eighth nerve, II for the proximal eighth nerve, III for the cochlear nucleus, IV for the olivary complex, and V for the lateral lemniscus (Fig. 10).
V III
Left 80 dB
.25 µV V
III A2
Left 70 dB
.25 µV
III I A3
109
V
Right 80 dB
.25 µV V I
A4
III Right 70 dB
.25 µV
Latency 1.00 ms/div Latency Offset –.80 ms
Figure 10 ABR of same patient with left intracanalicular acoustic neuroma. Click stimulus was performed at 70 dB and 80 dB for each ear. Note that latency increases with lower intensities. In this patient, the absolute latency of wave V is 5.96 for the left ear and 5.84 for the right ear (within normal limits). The interpeak latencies I to V are 4.00 for the left and 4.04 for the right (within normal limits).
110
Gidley
Table 8 Probability of an Acoustic Neuroma Based on Symptoms and Signs Probability Low probability (<5% chance)
Moderate probability (5–30% chance) High probability (>30% chance)
Factors Isolated vertigo Symmetrical hearing loss Historically explainable unilateral hypoacusis or tinnitus Sudden SNHL Unexplained tinnitus Combination of “classic symptoms and findings:” unilateral SNHL tinnitus decreased speech discrimination
Source: Refs. 62,75,96.
Acoustic Neuroma ABR’s place in the diagnosis of acoustic tumors still provokes debate among neuro-otologists. While some use ABR to screen all patients with asymmetric hearing loss, others reserve ABR for patients who have only a low probability of tumor (Table 8) (62,75,96), preferring to use MRI for patients who have a higher probability of having a tumor (96). Certainly, one could use MRI for all suspicious cases if availability and costs were not considerations, yet this is not the case in today’s healthcare environment. These arguments are largely based on the sensitivity of ABR to diagnose an AN. In large retrospective series of tumor patients, normal ABR can be seen in 2% to 18% of patients (105,109,110). In their series of 309 CPA tumors, Marangos et al. (110) found normal ABR in 50/261 sporadic ANs, 3/29 ANs due to NF-2, and 4/17 meningiomas. Their study further demonstrates quite elegantly the impact of MRI and the steady decrease in average tumor size by year of initial diagnosis. In their study population, in 1986 the average tumor size was 36 mm (+/– 10 mm); while in 1999, the average decreased to 16 mm (+/– 5 mm). ABR as a screening tool has been examined in prospective studies. Ruckenstien et al. (97) performed a prospective study of patients with asymmetric hearing loss examined with both ABR and MRI to determine the sensitivity, specificity, positive predictive value, and negative predictive value of ABR. Their small study included only 47 patients who met the inclusion criteria. Eight patients had significant pathology on MRI, five of which were ANs. However, ABR was normal in three of these patients; and all three had tumors smaller than 1.5 cm. Fourteen patients had an abnormality in ABR testing that was not demonstrable by MRI. In this preliminary study, the sensitivity of ABR to diagnose significant retrocochlear pathology was 63%, its specificity was 64%, the positive predictive value was 26% for AN, and its negative predictive value was 89% (97). Similar results were published by Ferguson et al. (106), though their protocol used contrast enhanced CT as a prerequisite prior to MRI imaging. Cueva (111) published a follow-up study to that of Ruckenstein et al. (97) and examined 312 adult patients with asymmetric hearing loss with ABR and MRI prospectively. He found 31 patients with retrocochlear pathology by MRI: 24 ANs and a collection of other pathologies (including 2 glomus jugulares and 1 petrous apex cholesterol granuloma). The ABR was abnormal in 22 of these 31 patients, thus nine patients had normal ABR, seven of which had an AN and all seven tumors were 16 mm or smaller. It should be noted that tumors as small as 5 mm were the cause of an abnormal ABR by these criteria. By his calculations, the sensitivity for ABR
was 71% and specificity was 74%. He recommends use of MRI for patients with asymmetric hearing loss, noting that if ABR is relied on as a screening test, 29 patients of 1000 screened will be missed (111). Overall sensitivity rates of ANs by ABR range from 88% to 95% (62,97,109,112–116). There are several reasons to account for such a wide range of positive findings: (i) testing parameters, (ii) tumor size, (iii) tumor location (intracanalicular versus extracanalicular), and (iv) involved nerve (superior versus inferior vestibular nerve). Waveform latencies and waveform morphologies have been studied by various authors. The most common indicators of pathology have been (i) an interaural wave I to V latency difference (IT5 or ILD I–V) >0.2 msec, (ii) a wave I–V interpeak latency (I–V IPL) >4.4 msec, and/or (iii) poor waveform morphology with either absent wave or no response. Absolute latencies are not considered as useful as interpeak latencies for the diagnosis of ANs because the absolute latency is affected by many factors, such as click intensity, hearing loss, and age (117). Some authors have not used interpeak latencies since wave I or II is difficult to identify even in normal listeners (112). Additionally, waveform amplitude is not used as a criterion since it is highly variable (112). A compilation of criteria used by several authors is presented in Table 9. Many different studies have examined the sensitivity of ABR by tumor size (Table 10), and a statistically significant positive correlation between tumor size and wave V latency has been reported (92). Additionally, when ABR waveforms from the contralateral ear are abnormal (e.g., a delayed wave V or prolonged wave I–V interval), a tumor larger than 2 cm should be suspected (118). Thus, it is concluded that ABR is nearly 100% sensitive for tumors larger than 2 cm (62,116,119). For this reason, ABR is a desirable screening test in the elderly and poor surgical risk patients for whom surgery may be indicated only for a symptomatic, large tumor (61,111). However, to have a reasonable chance of hearing preservation, tumors should be diagnosed as early (small) as possible, where small is defined as 2 cm or smaller (99,120). ABR sensitivity rates for tumors smaller than 1 cm range from 63% to 93% (62,97,113–116). Healthy patients with unilateral symptoms (hearing loss, poor discrimination, and tinnitus) should have MRI with gadolinium enhancement to find an AN. ABR still provides excellent insight into the physiology of the acoustic nerve, and this may have important implications for hearing preservation. Matthies and Samii (121) found that preoperative ABR was more important than the preoperative hearing quality for the chances of hearing preservation. They found that the presence of wave III correlated with better postoperative results, especially SDS (121). Robinette et al. (99) found similar results regarding the presence of waves I, III, and V when looking at preoperative predictors of hearing preservation. They reported that when these three waves are present, 61% of patients had hearing preservation, while only 27% of patients with one or more waves absent had hearing preserved (99). One can find papers that dispute any relationship between ABR waveforms and hearing preservation as well (120,122,123). It should be noted that poor ABR waveforms should not be used as a criteria to exclude the possibility of hearing preservation. Stidham and Roberson (124) reported a series of 30 patients undergoing middle fossa craniotomy for hearing preservation. They described seven patients with hearing improvement, classified as an increase in PTA ≥ 5 dB and/or an improvement in SDS by ≥ 12%. Interestingly, no patient with normal preoperative ABR experienced a hearing
Chapter 5: Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base
111
Table 9 Definitions of Abnormal ABR Interaural wave V latency difference, msec (ILD-V) (a.k.a. IT5)
Authors
Absolute wave V latency, msec
Interaural latency difference of I–V, msec (ILD I–V)
I–V interpeak latency, msec (I–V IPL)
Waveform morphology
House and Brackmann, 1979 (125) Bauch, 1982 (112)
>0.2
>6
Absence of wave V
>0.2
>6.1
Josey, 1988 (169)
>0.4
No response or poor overall waveform at high intensities Absence of V despite good hearing
Weiss, 1990 (170) Wilson, 1992 (113)
≥0.4 ≥0.4
>6.03
Dornhoffer, 1994 (114) Chandrasekhar, 1995 (115)
≥0.4 >0.2
>5.9
Gordon and Cohen, 1995 (62) Berrettini 1996 (83)
Ferguson, 1996 (106) Ruckenstein, 1996 (97)
>0.2
Saleh 1996 (95) Zappia, 1997 (116)
>0.3 >0.2
Godey, 1998 (84)
>0.2
Noguchi, 1999 (171) El Kashlan, 2000 (61)
≥0.3 >0.4
Haapaniemi, 2000 (144) Marangos, 2001 (110) Rupa, 2003 (119)
≥0.4 >0.3 ≥0.3
Cueva, 2004 (111)
>0.2
>4.4 ≥0.4
“abnormally prolonged” (>2 SD above normal limit for patient’s age and gender) >6.10 (male) >5.97 (female) Abnormal absolute wave V latency >6
≥4.45 ≥4.4
Abnormal ipsilateral or contralateral waveforms Abnormal or absent waveform morphology
>0.2 >0.3
>4.3
>0.3
>4.58 (male) >4.34 (female) ≥4.4
>0.2
>4.4 ≥4.4 >4.4
>7.75 >=0.4 >0.2
≥4.4 >4.4 ≥4.4
Abnormal absolute wave V latency
improvement. Of course, ABR is the most common technique used to monitor hearing intraoperatively (Chap. 6).
Other Tumors ABR results for posterior fossa meningiomas have similar rates of sensitivity as is found for ANs. In the pre-MRI era, House and Brackmann (125) found that only 75% of patients
Poor morphology in spite of adequate hearing
Absent or poor waveform morphology Absent waves, if adequate PTA Absent or abnormal waveform morphology Absent or abnormal waveforms Complete absence of waves if adequate PTA or absence of waves beyond wave I Abnormal or absent Absence of one or more waves, poor waveform morphology Absent or distorted waveform morphology
with non-AN pathology had abnormal ABRs. In their paper, 3 out of 10 meningiomas had normal ABR (125). Laird et al. (126) and Granick et al. (127) each found six out of six posterior fossa meningiomas had abnormal ABR. Aiba et al. (128) reported abnormal ABR in 8/10 cases; and Hart and Lillehie (129) reported abnormal ABR in 5/7 cases. Baguley contributed another 25 cases of CPA meningiomas, and found
Table 10 ABR Sensitivity with Respect to Tumor Size Study
No. of patients
Gordon and Cohen, 1995 (62) Chandrasekhar, 1995 (115) Bauch, 1996 Zappia, 1997 (116)
105 197
Marangos, 2001 (110)
309
Wilson, 1992 (113) Godey, 1998 (84)
111
40 89
≤10 mm
11–20 mm
>21 mm
69% 83% 82% 89% <15 mm 59.3%
88% 97%
100% 100%
98% 16–25 mm 82.7%
100% >25 mm 96.7%
Intracanalicular 66% 77%
Extracanalicular 96% 94%
112
Gidley
abnormal ABR in 100% of their tumors (100). Marangos et al. (110) found that 23.5% of meningiomas had normal ABR. Clearly, MRI is required to make the diagnosis of this tumor as well. Epidermoid tumors generally present at an advanced stage with multiple cranial nerve deficits and cerebellar signs (102,103). In the series by Quaranta (102), tumors ranged from 3.5 to 7 cm in maximum diameter. They found that ABR was normal in just one case. Absent or delayed waves were present in five cases, ipsilateral to the tumor. Four cases had bilateral abnormalities on ABR. Thus, 90% of their patients had abnormalities on ABR (102).
Stacked ABR Don et al. (130) described a new technique of ABR they called “stacked ABR.” In this test, ABR is obtained using 63 dB normal hearing level clicks in a high-pass noise-making procedure. The wave V amplitude is constructed by temporally aligning wave V of each derived-band ABR and summing the time-shifted responses. Using this technique, they found significantly lower wave V amplitudes in five AN patients who were missed by conventional ABR technique. These five tumors were all less than 1 cm in greatest dimension. They propose this technique as a cost-effective approach for AN screening. In a further study of stacked ABR, Philibert et al. (131) noticed that stacked ABR required a masking technique that might not be readily available. Additionally, they found that the relatively high intensity of the test might be annoying to the patient. Instead, they propose using tone burst to obtain a frequency-specific ABR.
Otoacoustic Emissions The phenomenon of sound being produced by the ear was first described in 1948 (132), and the definitive paper on otoacoustic emissions (OAEs) was published in 1978 (133). However, it was not until the 1990s that OAE testing became clinically widespread (134). OAE testing has enjoyed a significant increase in usage as part of a neonatal hearing screening strategy, monitoring ototoxicity or noise-induced hearing loss, and in suspected cases of functional hearing loss (134). OAEs are generated by outer hair cells of the cochlea (135). While OAEs are not useful as a screening test for ANs or other skull base tumors, they are measures of “cochlear reserve” and have been examined as possible predictors of hearing preservation (77,99,136). OAEs are divided into two groups: spontaneous and evoked. Spontaneous emissions are present in roughly 60% of normal ears (137). Evoked emissions are present in virtually 100% of normal ears (134). Evoked emissions are divided into distortion product (DPOAEs) and transient evoked (TEOAEs) emissions [Fig. 11(A) and 11(B)]. In a literature review, Robinette et al. (99) examined five studies describing 236 AN patients and reported that TEOAEs were present in at least one frequency in 47% of tumor ears. Brackmann et al. (77) described 333 AN patients considered for hearing preservation, 56 of these patients had DPOAEs measured. Normal DPOAEs were found in 91% (77). Ferber-Viart et al. (138) examined 168 AN patients with TEOAEs; (21%) had normal preoperative TEOAEs. They did not find an association with tumor size, functional symptoms, PTA, ABR, or electronystagmography (ENG) response with TEOAEs. Patients with TEOAEs tended to be younger, on average six years younger than those without TEOAEs.
In the subpopulation of 63 patients who underwent hearing conservation surgery, TEOAEs were present in 28% and absent in 72%. Sixty-six percent of those with TEOAEs present had hearing preserved, while only 44% with absent TEOAEs had hearing preserved. This difference was not statistically significant (138). In a more recent study, Kim et al. (136) examined 93 patients with AN that were candidates for hearing preservation. Fifty-one patients had hearing preserved. Eleven (22%) of these 51 patients had TEOAEs present in all five frequencies tested (1–4 kHz), while 40 (78%) had TEOAE responses anywhere from 0 to 4 of the frequencies tested. In the 42 patients who did not have hearing preserved, only three (7%) had positive TEOAEs in all five frequency bands, and 39 (93%) had TEOAEs from 0 to 4 frequency bands (p < 0.05). Other positive factors for hearing preservation in their series were small tumor size, tumor within the IAC, better hearing, and shorter latencies on ABR. Their conclusion was that a robust preoperative TEOAE pattern may be used as a favorable indicator for hearing preservation, especially when combined with the other positive factors listed above (136).
Electronystagmography Since its introduction in the 1960s, ENG (or more commonly now videonystagmography) has established itself as the most common test performed in evaluating patients with complaints of dizziness and vertigo (139). This test combines positional testing, optokinetic testing, random saccades and visual pursuit tests, and caloric stimulation to evaluate the vestibular ocular reflex and visual tracking centers of the brain. Findings on ENG for peripheral lesions are well described (140). ENG is an extremely valuable tool for examining the anatomic and functional integrity of the central and peripheral vestibular systems (141). The sensitivity of caloric testing for acoustic tumors ranges from 44% to 95% (83,140,142–144). Reduced or absent caloric response is the most frequent finding in AN patients (140,145,146). The amount of caloric weakness is proportional to the size of the tumor (140,145,146), although this has been disputed by others (92). Different authors use various criteria to describe a significant weakness; this can range from 20% (84) to 25% (77,147) and have a significant impact on the sensitivity and specificity of the test. The incidence of diminished caloric response by ENG for AN patients is presented in Table 11. More recently, head-shaking nystagmus (HSN) has been studied as a possible screening test for ANs. Humphriss et al. (27) studied 102 AN patients seen preoperatively. They used a passive head-shaking maneuver (1–2 Hz) and recorded eye movements with an ENG system. A significant response was five or more beats of nystagmus with a slow phase of at least 3 degrees/sec. In their study, significant caloric paresis was ≥ 25%. All patients had tumors confirmed by MRI and surgery. They found HSN in only 22 patients (i.e., sensitivity 22%). HSN was contralaterally beating in 19 patients (86%), ipsilaterally beating in three patients (14%), but absent in the remaining 80%. HSN was found more often in patients with either a greater canal paresis or have central vestibular signs than patients without HSN; however, the sensitivity still remains low [only 36% sensitivity even with severe (75–100%) canal paresis] (27). A low sensitivity rate (47.6%) for HSN was also reported by Asawavichiangianda et al. (148) for ANs. The range of reported sensitivity and specificity of HSN has been tabulated by Humphriss et al. (27). Sensitivity ranges from 22% to 95%; and specificity for a unilateral vestibular
Chapter 5: Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base
(A)
30
EAR: TONE PAIR:
SPL (dB)
20
113
Left Sequential
10 0 −10 −20
500
1K
8K
Score = (None)
Test result = (n/a) (B)
4K
2K F2 (Hz)
30
EAR: TONE PAIR:
SPL (dB)
20
Right: Sequential
10 0 −10 −20
500
1K
2K
4K
8K
F2 (Hz) Test result = (n/a)
Score = (None)
Figure 11 (A) Distortion product otoacoustic emissions (marked with Xs) from left ear of same patient with ipsilateral intracanalicular acoustic neuroma. Emissions are diminished but present from 1 to 3 kHz, and absent for higher frequencies, consistent with the audiometric findings. (B) Distortion product otoacoustic emissions (marked by circles) from right ear from same patient with left intracanalicular acoustic neuroma and are normal from 1 to 4 kHz.
disorder ranges from 53% to 92%. Across the 10 studies reviewed, many different criteria are used for the “gold standard” with ENG canal paresis definitions ranging from greater than 13% to greater than 30% difference. Additionally, the methods (active vs. passive) and nature of HSN (e.g., >3 beats, >5 beats, >2.5 degrees/sec, >6 degrees/sec) differed
across these different studies, which points out the variable nature of definitions used for HSN. Optokinetic and smooth pursuit abnormalities, when present, are reliable signs of brainstem compression (140). Berrettini (83) found a higher frequency of central findings in tumors greater than 3 cm.
Table 11 Incidence of Diminished Caloric Response by ENG Study Linthicum, 1979 (142) Haapaniemi, 2000 (144) Berrettini, 1996 (83) Naessens, 1996 (143) Godey, 1998 (84)
Small 43% Intracanalicular 55% Small tumors (<1 cm) 4/5 Overall 62.5% Overall 86%
Medium (1.1–3 cm) 16/18
Large 95% Extracanalicular 67% Large tumor (>3 cm) 16/16
114
Gidley
Caloric testing has been examined as a predictor of hearing preservation. The horizontal semicircular canal is stimulated through caloric testing, and thus an insight is gained regarding the superior vestibular nerve (144). Small superior vestibular nerve tumors have a more favorable prognosis for hearing preservation (77,113,149). Thus, it is reasoned that patients with reduced or absent caloric responses have a better chance of hearing preservation, since the superior vestibular nerve is involved. However, in practice, caloric results are not so clear cut. Linthicum (142) showed that 97.2% of superior nerve tumors had a caloric weakness, while only 60% of inferior vestibular nerve tumors had a caloric weakness. Holsinger et al. (150) examined 47 AN cases with planned hearing preservation. Their overall rate of measurable hearing postoperatively was 60%. ENG was obtained on 36 patients. Twenty-five patients demonstrated a significant unilateral weakness and measurable hearing was preserved in 14 (56%). Eleven patients had no caloric weakness and five had hearing preservation (45%). Brackmann et al. (77) published their series of 333 patients with tumors less than 2 cm considered candidates for hearing preservation. ENG was performed in 261 patients: 49% had “normal reduced response” (i.e., ≤25% weakness) and 51% had a reduced vestibular response (i.e., >25% weakness). They found similar rates of reduced responses across all hearing categories and that no significant difference existed between the preserved hearing groups and the no measurable hearing group (77). Despite its lack of sensitivity and its inability to discern superior from inferior nerve of origin, ENG might be helpful in identifying patients preoperatively that will have prolonged imbalance postoperatively. Driscoll et al. (151) found that central signs seen on ENG in AN patients portended a higher incidence of persistent (>3 mo) disequilibrium than those without central signs. Age greater than 55.5 years, female gender, constant preoperative disequilibrium present for >3.5 months were also associated with prolonged postoperative disequilibrium in their study.
Other Tumors The literature regarding ENG findings in meningioma is scant compared with that for ANs. Baguley et al. (100) compared the results of 18 of their patients with the tabulated results of caloric testing performed in four previous studies. Overall, abnormal caloric results were found in 55/67 (82%) patients; and the incidence of abnormal ABR ranged from 66% to 95% among the five studies. Marangos et al. (110) examined 309 CPA tumors, including 17 meningiomas, found on either CT or MRI. In the four meningiomas with normal ABR, ENG was normal in three. The article does not reveal the ENG results of the other 13 cases of meningiomas.
Rotatory Chair Testing Rotatory chair testing uses a computer controlled rotational stimulus of the horizontal VOR (usually from 0.01–0.64 Hz). Rotational stimuli produce a reflex, slow eye movement in the opposite direction of rotation with a rapid corrective saccade contralaterally (147). Gain, phase, asymmetry, and failure of visual fixation can be calculated at each test frequency and compared to age-specific normative data. The definition of abnormal findings varies from center to center, but a generally accepted rule is an abnormality of gain, phase, or symmetry seen in two frequencies (147). As with most balance tests, a “gold standard” is lacking with which to compare sensitivity and specificity. Most
Table 12 Test Conditions in Computerized Dynamic Posturography Sensory Organization Test Test 1 2 3 4 5 6
Condition Eyes open, fixed support Eyes closed, fixed support Visual surround referenced to sway, fixed support Eyes open, force plate referenced to sway Eyes closed, force plate referenced to sway Force plate and visual surround referenced to sway
reports on dizziness evaluation rely on history and physical findings to indicate “normal” and “abnormal”. Since the examining physicians are also the interpreting physicians of the balance test, there are no blinded comparison reviews. Rotary chair has been compared with ENG for sensitivity and specificity in identifying patients with vestibulopathy. In this regard, rotary chair has a higher sensitivity for peripheral vestibular pathology than ENG, but the specificity of ENG is higher than rotary chair (147).
Computerized Dynamic Posturography Computerized dynamic posturography (CDP) has been clinically available since 1986 (152). This test has two distinct parts: motor control test (MCT) and somatosensory organization test (SOT). MCT is a technique used to measure the functional ability of the subject to create adequate motor responses to changes in the pitch plane (58). Three trials are made of small, medium, large forward and backward movement and the latency, amplitude and symmetry of neuromuscular response to this movement (152). Electromyography (EMG) and biomechanical measures can be used to track the patient’s responses to each balance perturbation. The commercially available Neurocom MCT uses a strain gauge in the support surface, and not EMG, to track ankle-torque corrections during balance-correction. Normal latencies range between 130 and 160 msec for medium and large perturbations. The SOT measures the relative contributions from the vestibular, proprioceptive, and visual systems to maintain balance. This test uses a force platform, which can be stable or referenced to sway (move in a horizontal plane or pitch back and forth), and a visual surround, which can be stationary or referenced to sway, to determine the relative importance of visual, somatosensory, and vestibular input (153,154). Six different testing situations are created by combinations of stationary or moving force plate and stationary or moving visual surrounds (Table 12). The computer can change the position of the force plate or visual surround so that it remains in the same position relative to the patient’s sway (“sway referenced”). The interested reader is encouraged to read Allum and Sheperd’s (58) excellent review of this topic. The most commonly recognized pattern with vestibular lesions is abnormalities in situations five and six. Although CDP does not help in localizing a lesion, it does provide a functional measure of a patient’s ability to use properly the various input systems to maintain balance (153–155). Sensitivity of CDP to pick up vestibular pathology, in comparison to ENG, has been evaluated (141); however, these two modalities provide different types of information. Since not all imbalances are due to a vestibular lesion, CDP might certainly be abnormal in a patient with a normal ENG. Nonetheless, CDP gives information that is helpful regarding the functional status of the patient and might help to tailor a rehabilitation program for that patient (152).
Chapter 5: Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base
Levine et al. (156) used preoperative CDP to determine the nerve of origin for ANs less than 1.5 cm. In a small series, they found that patients with an inferior vestibular nerve tumor had abnormalities on SOT in conditions five and six, while patients whose tumors were from the superior vestibular nerve had normal CDP findings. Bergson and Sataloff (157) examined 21 patients with AN using CDP. They found abnormal test results in 81%, usually in conditions five and six. They found no correlation between the presence or severity of preoperative CDP results and postoperative balance function. Similar findings were reported by El-Kashlan et al. (155). Collins et al. (158) examined changes in balance following AN resection using balance posturography. They found patterns of abnormal sway and prolonged recovery times both pre- and postoperative, and these were most marked one month postoperatively. The limitations of CDP are that it does not provide lateralizing information or any information regarding cause (153). However, CDP does provide insight into how well patients can use their balance and how imbalance affects their activities of daily living.
Vestibular Evoked Potentials Vestibular evoked myogenic potentials (VEMPs) are shortlatency potentials recorded from surface electrodes over the tonically contracted SCM muscle evoked by high-level acoustic stimuli (159). The test subject is seated upright and asked to turn the head to the opposite side of the tested muscle. Surface electrodes are placed over the upper half of the SCM, while a ground electrode is placed on the forehead or sternum. Click stimuli sounds are delivered to the ipsilateral ear at intensities of 85 to 100 dB, and EMG is measured. The source of these responses is thought to be saccule. Matsuzaki et al. (160) described two patients with AN that had normal ABR results but abnormal VEMP. The tumors were 8 and 10 mm in size. They concluded that VEMP might be useful in the early diagnosis of ANs in patients with normal ABR. In a follow-up study, they review their experience with 87 AN patients, 79% of whom had decreased or absent VEMPs (161). Murofishi et al. (162) examined 21 patients with AN. They found abnormal or diminished VEMP ipsilateral to the tumor in 80% of patients, while all contralateral VEMPs were normal. Takeichi et al. (163) studied 18 patients with AN. They found diminished VEMP on the affected side in 13 patients (72%). They did not find any correlation with disequilibrium, spontaneous nystagmus, canal paresis, or pure-tone hearing. Tsutsumi et al. (164) examined 28 patients with AN and VEMP. They found no correlation between VEMP and caloric response, nerve of origin, audiometric threshold, or size of tumor. In a larger study of 170 patients with AN, Patko et al. (165), found abnormally low or absent VEMPs in 78.8% of patients. They did not find any correlation with horizontal canal weakness. Rauch has shown that VEMP has usefulness in the diagnosis of superior semicircular canal dehiscence syndrome and Meniere disease (166). Additionally, VEMP might provide some insight into brainstem pathology from stroke (167) or multiple sclerosis (168); however, at the time of this writing, their utility with AN or other skull base tumors is limited.
115
CONCLUSION The evaluation of patients with skull base tumors requires a rigorous history and physical examination. Careful evaluation of cranial nerve function is demanded. The examining physician should be aware of nontumor conditions that can mimic the findings of a skull base tumor. Judicious use of ancillary tests of neuro-otologic function helps in determining the extent of disease and plays an important role in deciding treatment methods. Some of these tests can be used to predict postoperative hearing and balance function.
REFERENCES 1. Bott L, Thumerelle C, Cuvellier JC, et al. Ataxia-telangiectasia: A review. Arch Pediatr. 2006;13(3):293–298. 2. Taylor AM, Byrd PJ. Molecular pathology of ataxia telangiectasia. J Clin Pathol. 2005;58(10):1009–1015. 3. Cross JH. Neurocutaneous syndromes and epilepsy—Issues in diagnosis and management. Epilepsia. 2005;46 Suppl 10:17–23. 4. Kreusel KM. Ophthalmological manifestations in VHL and NF 1: Pathological and diagnostic implications. Fam Cancer. 2005;4(1):43–47. 5. Patel NP, Mhatre AN, Lalwani AK. Molecular pathogenesis of skull base tumors. Otol Neurotol. 2004;25(4):636–643. 6. Xiao GH, Chernoff J, Testa JR. NF2: The wizardry of merlin. Genes Chromosomes Cancer. 2003;38(4):389–399. 7. Smith DS, Lee KK, Milczuk HA. Otolaryngologic manifestations of PHACE syndrome. Int J Pediatr Otorhinolaryngol. 2004;68(11):1445–1450. 8. Comi AM. Pathophysiology of Sturge-Weber syndrome. J Child Neurol. 2003;18(8):509–516. 9. Freedman SF, Amedee RG, Molony T. Neurotologic manifestations of von Hippel Lindau disease. Ear Nose Throat J. 1992;71(12):655–658. 10. Brandt T, Strupp M. General vestibular testing. Clin Neurophysiol. 2005;116(2):406–426. 11. Goldberg M. The control of gaze. In: Kandell E, Schwartz J, Jessell T, eds. Principles of Neural Science. New York: McGrawHill, 2000 782–800. 12. Glanze W, ed. Mosby’s Medical Dictionary. 3rd ed. St. Louis: C.V. Mosby, 1990. 13. Ross R. How to Examine: The Nervous System. New York: Medical Publishing Co., 1985. 14. Traccis S et al. Evaluating patients with vertigo: Bedside examination. Neurol Sci. 2004;25 Suppl 1:S16–S19. 15. Leigh R, Zee D. The Neurology of Eye Movements, 4th ed. New York: Oxford University Press, 2006. 16. Lee K. The vestibular system and its disorders. In: Lee K, ed. Essential Otolaryngology. New York: Medical Examination Publishing, 1995 93–124. 17. Nedzelski JM. Cerebellopontine angle tumors: Bilateral flocculus compression as cause of associated oculomotor abnormalities. Laryngoscope. 1983;93(10):1251–1260. 18. Longridge NS, Mallinson AI. The dynamic illegible E (DIE) test: A simple technique for assessing the ability of the vestibuloocular reflex to overcome vestibular pathology. J Otolaryngol. 1987;16(2):97–103. 19. Herdman SJ et al. Computerized dynamic visual acuity test in the assessment of vestibular deficits. Am J Otol. 1998;19(6):790– 796. 20. Dix MR, Hallpike CS. The pathology symptomatology and diagnosis of certain common disorders of the vestibular system. Proc R Soc Med. 1952;45(6):341–354. 21. Halmagyi GM, Curthoys IS. A clinical sign of canal paresis. Arch Neurol. 1988;45(7):737–739. 22. Schubert MC et al. Optimizing the sensitivity of the head thrust test for identifying vestibular hypofunction. Phys Ther. 2004;84(2):151–158.
116
Gidley
23. Foster CA et al. Functional loss of the horizontal doll’s eye reflex following unilateral vestibular lesions. Laryngoscope. 1994;104(4):473–478. 24. Harvey SA, Wood DJ. The oculocephalic response in the evaluation of the dizzy patient. Laryngoscope. 1996;106(1 Pt 1):6–9. 25. Harvey SA, Wood DJ, Feroah TR. Relationship of the head impulse test and head-shake nystagmus in reference to caloric testing. Am J Otol. 1997;18(2):207–213. 26. Beynon GJ, Jani P, Baguley DM. A clinical evaluation of head impulse testing. Clin Otolaryngol Allied Sci. 1998;23(2):117–122. 27. Humphriss RL, Baguley DM, Moffat DA. Head-shaking nystagmus in patients with a vestibular schwannoma. Clin Otolaryngol Allied Sci. 2003;28(6):514–519. 28. Shockley WW, Stucker FJ Jr. Squamous cell carcinoma of the external ear: A review of 75 cases. Otolaryngol Head Neck Surg. 1987;97(3):308–312. 29. Ross DA, Sasaki CT. Cancer of the Ear and Temporal Bone. In: Myers E., Suen J, ed. Cancer of the Head and Neck. Philadelphia, PA: W.B. Saunders, 1996; 586–597. 30. Clayman GL et al. Mortality risk from squamous cell skin cancer. J Clin Oncol. 2005;23(4):759–765. 31. Burkey JM et al. Clinical utility of the 512-Hz Rinne tuning fork test. Am J Otol. 1998;19(1):59–62. 32. Vikram KB, Naseeruddin K. Combined tuning fork tests in hearing loss: Explorative clinical study of the patterns. J Otolaryngol. 2004;33(4):227–234. 33. Poe DS, Abou-Halawa A, Abdel-Razek O. Analysis of the dysfunctional eustachian tube by video endoscopy. Otol Neurotol. 2001;22(5):590–595. 34. Meyerhoff WL, Anderson RG. Non-auditory presentations of cerebellopontine angle lesions. Laryngoscope. 1984;94(7):904– 906. 35. Nemzek WR. The trigeminal nerve. Top Magn Reson Imaging. 1996;8(3):132–154. 36. Saper C. Brain stem, reflexive behavior, and the cranial nerves. In: Kandell E, Schwartz J, Jessell T, eds. Principles of Neural Science. New York: McGraw-Hill, 2000 873–888. 37. Carpenter M. Core Text of Neuroanatomy. 3rd ed. Baltimore: Williams and Wilkins, 1985. 38. Doty RL, Mishra A. Olfaction and its alteration by nasal obstruction, rhinitis, and rhinosinusitis. Laryngoscope. 2001;111(3):409–423. 39. Rombaux P et al. Smell disorders in ENT clinic. B-Ent. 2005;Suppl 1:97–107; quiz 108–109. 40. Hummel T, Welge-Luessen A. Assessment of olfactory function. Adv Otorhinolaryngol. 2006;63:84–98. 41. Wrobel BB, Leopold DA. Clinical assessment of patients with smell and taste disorders. Otolaryngol Clin North Am. 2004;37(6):1127–1142. 42. Schiffman SS. Taste and smell in disease (first of two parts). N Engl J Med. 1983;308(21):1275–1279. 43. Doty RL et al. University of Pennsylvania Smell Identification Test: A rapid quantitative olfactory function test for the clinic. Laryngoscope. 1984;94(2 Pt 1):176–178. 44. Jackson CG et al. Facial paralysis of neoplastic origin: Diagnosis and management. Laryngoscope. 1980;90(10 Pt 1):1581–1595. 45. House JW, Brackmann DE. Facial nerve grading system. Otolaryngol Head Neck Surg. 1985;93(2):146–147. 46. Yeh S, Foroozan R. Orbital apex syndrome. Curr Opin Ophthalmol. 2004;15(6):490–498. 47. Zaffaroni M, Baldini SM, Ghezzi A. Cranial nerve, brainstem and cerebellar syndromes in the differential diagnosis of multiple sclerosis. Neurol Sci. 2001;22 Suppl 2:S74–S78. 48. Krasnianski M, Neudecker S, Zierz S. The Schmidt and Vernet classical syndrome. Alternating brain stem syndromes that do not exist?. Nervenarzt. 2003;74(12):1150–1154. 49. Johnston JL. Parasellar syndromes. Curr Neurol Neurosci Rep. 2002;2(5):423–431. 50. Schweinfurth JM, Johnson JT, Weissman J. Jugular foramen syndrome as a complication of metastatic melanoma. Am J Otolaryngol. 1993;14(3):168–174. 51. Paparounas K et al. Collet-Sicard syndrome disclosing glomus tumor of the skull base. Eur Neurol. 2003;49(2):103–105.
52. Chacon G, Alexandraki I, Palacio C. Collet-sicard syndrome: An uncommon manifestation of metastatic prostate cancer. South Med J. 2006;99(8):898–899. 53. Romberg M. Lehrbuch Der Nervenkrankheiten Des Menschen. Berlin: Duncker, 1846. 54. Weber PC, Cass SP. Clinical assessment of postural stability. Am J Otol. 1993;14(6):566–569. 55. Fukuda T. Statokinetic Reflexes in Equilibrium and Movement. Tokyo, Japan: University of Tokyo Press, 1983. 56. Hickey SA et al. Unterberger stepping test: A useful indicator of peripheral vestibular dysfunction? J Laryngol Otol. 1990;104(8):599–602. 57. Bonanni M, Newton R. Test-retest reliability of the Fukuda Stepping Test. Physiother Res Int. 1998;3(1):58–68. 58. Allum JH, Shepard NT. An overview of the clinical use of dynamic posturography in the differential diagnosis of balance disorders. J Vestib Res. 1999;9(4):223–252. 59. Kuipers-Upmeijer J, Oosterhuis HJ. Unterberger’s test not useful in testing of vestibular function. Ned Tijdschr Geneeskd. 1994;138(3):136–139. 60. Horak FB. Clinical measurement of postural control in adults. Phys Ther. 1987;67(12):1881–1885. 61. El-Kashlan HK, Eisenmann D, Kileny PR. Auditory brain stem response in small acoustic neuromas. Ear Hear. 2000;21(3):257– 262. 62. Gordon ML, Cohen NL. Efficacy of auditory brainstem response as a screening test for small acoustic neuromas. Am J Otol. 1995;16(2):136–139. 63. Cushing H. Tumors of the Nervus Acusticus and the Syndrome of the Cerebellopontine Angle. Philadelphia, PA: WB Saunders, 1917. 64. Johnson EW. Auditory test results in 110 surgically confirmed retrocochlear lesions. J Speech Hear Disord. 1965;30(4):307–317. 65. Glasscock ME 3rd, Levine SC, McKennan KX. The changing characteristics of acoustic neuroma patients over the last 10 years. Laryngoscope. 1987;97(10):1164–1167. 66. Jerger J. Clinical experience with impedance audiometry. Arch Otolaryngol. 1970;92(4):311–324. 67. Jerger J. Impedance terminology. Arch Otolaryngol. 1975;101(10):589–590. 68. Jerger J, Hayes D. Diagnostic speech audiometry. Arch Otolaryngol. 1977;103(4):216–222. 69. Hayes D, Jerger J. Impedance audiometry in otologic diagnosis. Otolaryngol Clin North Am. 1978;11(3):759–767. 70. Clemis JD, Mc Gee T. Brain stem electric response audiometry in the differential diagnosis of acoustic tumors. Laryngoscope. 1979;89(1):31–42. 71. Dix MR, Hallpike CS. The otoneurological diagnosis of tumours of the VIII nerve. Proc R Soc Med. 1958;51(11):889–896. 72. Walsh TE, Goodman A. Speech discrimination in central auditory lesions. Laryngoscope. 1955;65(1):1–8. 73. Hirsch A, Anderson H. Audiologic test results in 96 patients with tumours affecting the eighth nerve. A clinical study with emphasis on the early audiological diagnosis. Acta Otolaryngol Suppl. 1980;369:1–26. 74. Telischi F. An objective method of analyzing cochlear versus noncochlear patterns of distortion-product otoacoustic emissions in patients with acoustic neuromas. Laryngoscope. 2000;110(4):553–562. 75. Selesnick SH, Jackler RK. Clinical manifestations and audiologic diagnosis of acoustic neuromas. Otolaryngol Clin North Am. 1992;25(3):521–551. 76. Committee on Hearing and Equilibrium guidelines for the evaluation of hearing preservation in acoustic neuroma (vestibular schwannoma). American Academy of Otolaryngology-Head and Neck Surgery Foundation, INC. Otolaryngol Head Neck Surg. 1995;113(3):179–180. 77. Brackmann DE et al. Prognostic factors for hearing preservation in vestibular schwannoma surgery. Am J Otol. 2000;21(3):417– 24. 78. Anderson H, Barr B, Wedenberg E. The early detection of acoustic tumours by the stapedius reflex test. In: Sensorineural Hearing Loss. Ciba Foundation Symposium. 1970: p. 275–294.
Chapter 5: Head, Neck, and Neuro-otologic Assessment of Patients with Tumors of the Skull Base 79. Harner SG, Laws ER Jr. Clinical findings in patients with acoustic neurinoma. Mayo Clin Proc. 1983;58(11):721–728. 80. Moffat DA, Hardy DG. Early diagnosis and surgical management of acoustic neuroma: Is it cost effective? J R Soc Med. 1989;82(6):329–332. 81. Thomsen J, Tos M. Acoustic neuroma: Clinical aspects, audiovestibular assessment, diagnostic delay, and growth rate. Am J Otol. 1990;11(1):12–19. 82. Kanzaki J et al. Audiological findings in acoustic neuroma. Acta Otolaryngol Suppl. 1991;487:125–132. 83. Berrettini S et al. Acoustic neuroma: Correlations between morphology and otoneurological manifestations. J Neurol Sci. 1996;144(1-2):24–33. 84. Godey B et al. Sensitivity of auditory brainstem response in acoustic neuroma screening. Acta Otolaryngol. 1998;118(4):501– 504. 85. Moffat DA et al. Contralateral auditory brainstem response abnormalities in acoustic neuroma. J Laryngol Otol. 1989;103(9):835–838. 86. Beck HJ et al. Acoustic neuromas with normal pure tone hearing levels. Otolaryngol Head Neck Surg. 1986;94(1):96–103. 87. Roland PS et al. Normal hearing in patients with acoustic neuroma. South Med J. 1987;80(2):166–169. 88. Selesnick SH, Jackler RK, Pitts LW. The changing clinical presentation of acoustic tumors in the MRI era. Laryngoscope. 1993;103(4 Pt 1):431–436. 89. Acoustic Neuroma: NIH Consensus Statement. NIH, 1991:1–24. 90. Morrison A. Acoustic Neuroma. In: Management of Sensorineural Hearing Loss. London, England: Butterworths, 1975:46–79. 91. Telian SA, Kileny PR. Pitfalls in neurotologic diagnosis. Ear Hear. 1988;9(2):86–91. 92. Rosenberg SI. Natural history of acoustic neuromas. Laryngoscope. 2000;110(4):497–508. 93. Schuknecht H. Pathology of the Ear. Cambridge, MA: Harvard University Press, 1974. 94. Graamans K, Van Dijk JE, Janssen LW. Hearing deterioration in patients with a non-growing vestibular schwannoma. Acta Otolaryngol. 2003;123(1):51–54. 95. Saleh EA et al. Normal hearing in acoustic neuroma patients: A critical evaluation. Am J Otol. 1996;17(1):127–132. 96. Welling DB et al. Acoustic neuroma: A cost-effective approach. Otolaryngol Head Neck Surg. 1990;103(3):364–370. 97. Ruckenstein MJ et al. A prospective study of ABR and MRI in the screening for vestibular schwannomas. Am J Otol. 1996;17(2):317–320. 98. Obholzer RJ, Rea PA, Harcourt JP. Magnetic resonance imaging screening for vestibular schwannoma: Analysis of published protocols. J Laryngol Otol. 2004;118(5):329–332. 99. Robinette MS et al. Nonsurgical factors predictive of postoperative hearing for patients with vestibular schwannoma. Am J Otol. 1997;18(6):738-745. 100. Baguley DM et al. Audio-vestibular findings in meningioma of the cerebello-pontine angle: A retrospective review. J Laryngol Otol. 1997;111(11):1022–1026. 101. Doyle K, De la Cruz A. Cerebellopontine angle epidermoids: Results of surgical treatment. Skull Base Surgery. 1996;6:27– 33. 102. Quaranta N et al. Audiologic presentation of cerebellopontine angle cholesteatoma. J Otolaryngol. 2003;32(4):217–221. 103. Kaylie DM et al. Neurotologic management of intracranial epidermoid tumors. Laryngoscope. 2005;115(6):1082–1086. 104. Jewett DL, Williston JS. Auditory-evoked far fields averaged from the scalp of humans. Brain. 1971;94(4):681–696. 105. Brackmann DE, Selters WA. Brainstem electric audiometry: Acoustic neurinoma detection. Rev Laryngol Otol Rhinol (Bord). 1979;100(1-2):49–51. 106. Ferguson MA et al. Efficiency of tests used to screen for cerebello-pontine angle tumours: A prospective study. Br J Audiol. 1996;30(3):159–176. 107. Bu-Saba NY et al. Significance of false-positive auditory brainstem response: A clinical study. Am J Otol. 1994;15(2):233–236. 108. Hall J. Handbook of Auditory Evoked Responses. Boston, MA: Allyn and Bacon, 1992.
117
109. Josey AF, Jackson CG, Glasscock ME 3rd. Brainstem evoked response audiometry in confirmed eighth nerve tumors. Am J Otolaryngol. 1980;1(4):285–290. 110. Marangos N et al. Brainstem response in cerebellopontine angle tumors. Otol Neurotol. 2001;22(1):95–99. 111. Cueva RA. Auditory brainstem response versus magnetic resonance imaging for the evaluation of asymmetric sensorineural hearing loss. Laryngoscope. 2004;114(10):1686–1692. 112. Bauch CD, Rose DE, Harner SG. Auditory brain stem response results from 255 patients with suspected retrocochlear involvement. Ear Hear. 1982;3(2):83–86. 113. Wilson DF et al. The sensitivity of auditory brainstem response testing in small acoustic neuromas. Laryngoscope. 1992;102(9):961–964. 114. Dornhoffer JL, Helms J, Hoehmann DH. Presentation and diagnosis of small acoustic tumors. Otolaryngol Head Neck Surg. 1994;111(3 Pt 1):232–235. 115. Chandrasekhar SS, Brackmann DE, Devgan KK. Utility of auditory brainstem response audiometry in diagnosis of acoustic neuromas. Am J Otol. 1995;16(1):63–67. 116. Zappia JJ et al. Rethinking the use of auditory brainstem response in acoustic neuroma screening. Laryngoscope. 1997;107(10):1388–1392. 117. Burkey JM et al. Acoustic reflexes, auditory brainstem response, and MRI in the evaluation of acoustic neuromas. Laryngoscope. 1996;106(7):839–841. 118. Musiek FE, Kibbe K. Auditory brain stem response wave IVV abnormalities from the ear opposite large cerebellopontine lesions. Am J Otol. 1986;7(4):253–257. 119. Rupa V et al. Cost-effective initial screening for vestibular schwannoma: Auditory brainstem response or magnetic resonance imaging? Otolaryngol Head Neck Surg. 2003;128(6):823– 828. 120. Nadol JB Jr et al. Preservation of hearing and facial nerve function in resection of acoustic neuroma. Laryngoscope. 1992;102(10):1153–1158. 121. Matthies C, Samii M. Management of vestibular schwannomas (acoustic neuromas): The value of neurophysiology for intraoperative monitoring of auditory function in 200 cases. Neurosurgery. 1997;40(3):459–466; discussion 466–468. 122. Nadol JB Jr et al. Preservation of hearing in surgical removal of acoustic neuromas of the internal auditory canal and cerebellar pontine angle. Laryngoscope. 1987;97(11):1287–1294. 123. Kemink JL et al. Hearing preservation following suboccipital removal of acoustic neuromas. Laryngoscope. 1990;100(6):597– 602. 124. Stidham KR, Roberson JB Jr. Hearing improvement after middle fossa resection of vestibular schwannoma. Otol Neurotol. 2001;22(6):917–921. 125. House JW, Brackmann DE. Brainstem audiometry in neurotologic diagnosis. Arch Otolaryngol. 1979;105(6):305–309. 126. Laird FJ et al. Meningiomas of the cerebellopontine angle. Otolaryngol Head Neck Surg. 1985;93(2):163–167. 127. Granick MS et al. Cerebellopontine angle meningiomas: Clinical manifestations and diagnosis. Ann Otol Rhinol Laryngol. 1985;94(1 Pt 1):34–38. 128. Aiba T et al. Clinical characteristics of rare cerebellopontine angle tumours: Comparison with acoustic tumors. In: Proceedings of the First International Conference on Acoustic Neuroma. New York, NY: Kugler, 1992. 129. Hart MJ, Lillehei KO. Management of posterior cranial fossa meningiomas. Ann Otol Rhinol Laryngol. 1995;104(2):105– 116. 130. Don M et al. Successful detection of small acoustic tumors using the stacked derived-band auditory brain stem response amplitude. Am J Otol. 1997;18(5):608–621; discussion 682–685. 131. Philibert B et al. Stacked tone-burst-evoked auditory brainstem response (ABR): Preliminary findings. Int J Audiol. 2003;42(2):71–81. 132. Gold T. Hearing II. The physical basis of the action of the cochlea. Proc Royal Soc Britain. 1948;135:492–498. 133. Kemp DT. Stimulated acoustic emissions from within the human auditory system. J Acoust Soc Am. 1978;64(5):1386–1391.
118
Gidley
134. Hall J. Handbook of Otoacoustic Emissions. San Diego, CA: Singular, 2000. 135. Brownell WE et al. Evoked mechanical responses of isolated cochlear outer hair cells. Science. 1985;227(4683):194–196. 136. Kim AH et al. Transient evoked otoacoustic emissions pattern as a prognostic indicator for hearing preservation in acoustic neuroma surgery. Otol Neurotol. 2006;27(3):372–379. 137. Burns EM, Arehart KH, Campbell SL. Prevalence of spontaneous otoacoustic emissions in neonates. J Acoust Soc Am. 1992;91(3):1571–1575. 138. Ferber-Viart C et al. Is the presence of transient evoked otoacoustic emissions in ears with acoustic neuroma significant? Laryngoscope. 1998;108(4 Pt 1):605–609. 139. Jongkees LB, Philipszoon AJ. Electronystagmography. Acta Otolaryngol Suppl. 1964;189. 140. McGee ML. Electronystagmography in peripheral lesions. Ear Hear. 1986;7(3):167–175. 141. Amin M et al. A comparison of electronystagmography results with posturography findings from the BalanceTrak 500. Otol Neurotol. 2002;23(4):488–493. 142. Linthicum F, Khalessi M, Churchill D. Electronystagmographic caloric bithermal vestibular test (ENG): results in acoustic tumor cases. In: House and Luetje C eds. Acoustic Tumors. Baltimore: University Park Press, 1979 pp 237–240. 143. Naessens B et al. Re-evaluation of the ABR in the diagnosis of CPA tumors in the MRI-era. Acta Otorhinolaryngol Belg. 1996;50(2):99–102. 144. Haapaniemi JJ et al. Audiovestibular findings and location of an acoustic neuroma. Eur Arch Otorhinolaryngol. 2000;257(5):237– 241. 145. Bergenius J, Magnusson M. The relationship between caloric response, oculomotor dysfunction and size of cerebello-pontine angle tumours. Acta Otolaryngol. 1988;106(5-6):361–367. 146. Guyot JP et al. Diagnosis of cerebellopontine angle tumors. ORL J Otorhinolaryngol Relat Spec. 1992;54(3):139–143. 147. Arriaga MA, Chen DA, Cenci KA. Rotational chair (ROTO) instead of electronystagmography (ENG) as the primary vestibular test. Otolaryngol Head Neck Surg. 2005;133(3):329–333. 148. Asawavichiangianda S et al. Significance of head-shaking nystagmus in the evaluation of the dizzy patient. Acta Otolaryngol Suppl. 1999;540:27–33. 149. Shelton C et al. Acoustic tumor surgery. Prognostic factors in hearing conversation. Arch Otolaryngol Head Neck Surg. 1989;115(10):1213–1216. 150. Holsinger FC, Coker NJ, Jenkins HA. Hearing preservation in conservation surgery for vestibular schwannoma. Am J Otol. 2000;21(5):695–700. 151. Driscoll CL et al. Preoperative identification of patients at risk of developing persistent dysequilibrium after acoustic neuroma removal. Am J Otol. 1998;19(4):491–495. 152. Voorhees RL. The role of dynamic posturography in neurotologic diagnosis. Laryngoscope. 1989;99(10 Pt 1):995–1001. 153. Furman JM. Role of posturography in the management of vestibular patients. Otolaryngol Head Neck Surg. 1995;112(1):8–15.
154. Monsell EM et al. Computerized dynamic platform posturography. Otolaryngol Head Neck Surg. 1997;117(4):394– 398. 155. El-Kashlan HK et al. Disability from vestibular symptoms after acoustic neuroma resection. Am J Otol. 1998;19(1):104– 111. 156. Levine SC, Muckle RP, Anderson JH. Evaluation of patients with acoustic neuroma with dynamic posturography. Otolaryngol Head Neck Surg. 1993;109(3 Pt 1):392–398. 157. Bergson E, Sataloff RT. Preoperative computerized dynamic posturography as a prognostic indicator of balance function in patients with acoustic neuroma. Ear Nose Throat J. 2005;84(3):154–156. 158. Collins MM et al. Dynamic assessment of imbalance in acoustic neuroma patients by sway magnetometry. Clin Otolaryngol. 2000;25(6):570–576. 159. Akin FW, Murnane OD. Vestibular evoked myogenic potentials: Preliminary report. J Am Acad Audiol. 2001;12(9):445-452; quiz 491. 160. Matsuzaki M, Murofushi T, Mizuno M. Vestibular evoked myogenic potentials in acoustic tumor patients with normal auditory brainstem responses. Eur Arch Otorhinolaryngol. 1999;256(1):1– 4. 161. Ushio M et al. Click- and short tone burst-evoked myogenic potentials in cerebellopontine angle tumors. Acta Otolaryngol Suppl. 2001;545:133–135. 162. Murofushi T, Matsuzaki M, Mizuno M. Vestibular evoked myogenic potentials in patients with acoustic neuromas. Arch Otolaryngol Head Neck Surg. 1998;124(5):509–512. 163. Takeichi N et al. Vestibular evoked myogenic potential (VEMP) in patients with acoustic neuromas. Auris Nasus Larynx. 200128 Suppl:S39–S41. 164. Tsutsumi T et al. Prediction of the nerves of origin of vestibular schwannomas with vestibular evoked myogenic potentials. Am J Otol. 2000;21(5):712–715. 165. Patko T et al. Vestibular evoked myogenic potentials in patients suffering from an unilateral acoustic neuroma: A study of 170 patients. Clin Neurophysiol. 2003;114(7):1344–1350. 166. Rauch SD. Vestibular evoked myogenic potentials. Curr Opin Otolaryngol Head Neck Surg. 200614(5):299–304. 167. Chen CH, Young YH. Vestibular evoked myogenic potentials in brainstem stroke. Laryngoscope. 2003;113(6):990–993. 168. Alpini D et al. Vestibular evoked myogenic potentials in multiple sclerosis: Clinical and imaging correlations. Mult Scler. 2004;10(3):316–321. 169. Josey AF, Glasscock ME 3rd, Musiek FE. Correlation of ABR and medical imaging in patients with cerebellopontine angle tumors. Am J Otol. 1988;9 Suppl:12–16. 170. Weiss MH, Kisiel DL, Bhatia P. Predictive value of brainstem evoked response in the diagnosis of acoustic neuroma. Otolaryngol Head Neck Surg. 1990;103(4):583–585. 171. Noguchi Y, Komatsuzaki A, Nishida H. Cochlear microphonics for hearing preservation in vestibular schwannoma surgery. Laryngoscope. 1999;109(12):1982–1987.
6 Anesthesia and Intraoperative Monitoring of Patients with Tumors of the Skull Base Walter S. Jellish and Steven B. Edelstein
INTRODUCTION
for posterior fossa surgeries, with some modifications. The prone position is particularly useful for accessing lesions at or near midline and the fourth ventricle (3). Prone positioning is associated with numerous hemodynamic and respiratory changes that must be monitored closely. Significant V/Q mismatching is present in the prone position and access to the airway is compromised. Vigilance by the anesthesiologist is required when positioning the patient since dislodging of the endotracheal tube will require emergency re-intubation in suboptimal conditions. When transferring patients from supine to prone, significant hemodynamic changes can occur, secondarily to acute changes in preload from either abdominal compression or thoracic impedance. Cardiac dysrhythmias may occur from changes in preload as a result of this compression, thus electrocardiographic monitoring is essential (4). Several items, such as chest rolls and orthopedic frames, have been developed to help decrease these side effects. Some of the frames include: the Jackson spine table, simple chest and abdominal rolls composed of fabric or gel, and padded square frames that have a large opening for the abdominal contents. Each frame or support device is associated with its own series of risks and benefits that is beyond the scope of this article. Other than hemodynamic and respiratory problems associated with the prone position, if a slight head-up position is used, the patient is at risk for venous air embolism (VAE). VAE is a significant concern during skull base procedures and this will be discussed in further detail later in this chapter. Prolonged prone positioning has also been associated with significant facial edema, orbital/facial swelling, central venous retinal thrombosis, and posterior ischemic optic neuropathy—a disastrous condition that can result in permanent blindness.
Many issues surround the administration of anesthesia for patients undergoing surgery of the skull base. Not only are there the difficulties surrounding exposure of deeply seated anatomic structures, but there are also issues regarding positioning, prevention of iatrogenic nerve injury, and blood loss. This chapter describes many of the issues surrounding the delivery of anesthesia for skull base surgery and will touch on some of the neurophysiological monitors that can be used to improve the outcome. The review of this subjective matter is by no means exhaustive, but is meant to give some insight into the multiple and complex problems faced by the operative team during these procedures.
POSITIONING One of the challenges of skull base surgery has to do with positioning. The exact approach will depend on patient’s anatomy, clinical status, and tumor size. These approaches relate to the bone routes the surgeon will take in order to reach the neoplasm. These approaches include craniofacial, orbitocranial, infratemporal, and suboccipital (transcondylar). Lateral approaches include retrosigmoid, translabyrinthine, and orbitocranial zygomatic (1). Each of these approaches is associated with particular positions that have specific morbidities. Some of these morbidities and concerns will be discussed in further detail.
Supine Position Anterior skull base lesions can be quite difficult to surgically approach, though the physical position of the patient is that of supine. Supine position is, for the anesthesiologist, the easiest one to manage. Access to the airway is simple, though the patient may be rotated 180 degrees. There are limited changes in the patient’s hemodynamic profile, but there are significant changes in the pulmonary system, especially those related to diaphragmatic elevation. This altered diaphragmatic elevation ultimately promotes atelectasis and ventilation-to-perfusion (V/Q) mismatching. Posterior fossa lesions can also be accessed via the supine position, but require the head be laterally rotated and flexed. This flexion is sometimes impossible, especially in the elderly population and may be associated with venous obstruction (2).
Park-Bench Position The park-bench position (lateral oblique position) is commonly used in skull base procedures involving the posterior fossa. It is a semiprone lateral position with the head flexed and slightly elevated—5 to 10 degrees. It is quicker than the true prone and allows both lateral and midline approaches to the posterior fossa (3). The flexion of the neck may impinge on the venous circulation and an obstruction may not always be recognized by examination of the external position (5). This position has less hemodynamic and respiratory effects than the full prone position, but the neck flexion may be associated with brachial plexus injuries if care is not taken when final position is achieved. Extreme flexion may compromise spinal cord perfusion and has been associated with quadriplegia (5).
Prone Position The prone position is one of the most frequently used positions for spine surgery, though it is sometimes used 119
120
Jellish and Edelstein
Lateral Position This position is frequently used for intracerebellar procedures involving lateral or cerebellar hemispheric lesions, lesions of the clivus, petrous ridge, anterior and lateral foramen magnum. Typically the retromastoid, transtentorial, and transcondylar surgical approaches require this position. Again, there are some significant hemodynamic and pulmonary implications to the position. Significant V/Q mismatching takes place, though to a much less degree than the prone position. Some of the disadvantages of this position include the potential for lateral popliteal nerve palsy in the dependent leg, compression of the inferior shoulder and axillary structures, and the superior shoulder being in the line of sight of the surgeon (3). As such, it is important that the anesthesiologist confirms that the final position of the patient is appropriate. Devices such as axillary rolls (designed to decompress dependent axillary neurovascular structures) must be correctly placed and leg positions be padded.
Sitting Position Of all positions used by surgeons involved with skull base procedures, the sitting position has undergone the most scrutiny. The theoretical advantage of the sitting position is that it allows for improved cerebral relaxation and promotes gravity drainage of blood and cerebral spinal fluid (7). However, the complications are numerous and include hemodynamic instability, VAE with the possibility of paradoxical air embolism, pneumocephalus, quadriplegia (especially in the presence of extreme neck flexion), and compressive neuropathy (8,9). Another potential complication includes inadequate cerebral perfusion after assuming the sitting position, but may be balanced by a reduction in cerebral rate of metabolic oxygen consumption (10). Early physiological studies reveal significant changes when a patient assumes the upright position. A change in cardiac and systemic vascular resistance is known to increase 19% and 10%, respectively. In addition, stroke volume and cardiac index may decrease as much as 21% and 10%, respectively (11). In the presence of inhalational anesthetic agents, arterial hypotension may be profound, especially since these agents may cause vasodilation, venous pooling in the lower extremities, and dose-dependent cardiovascular depression (12). It is essential that the anesthesiologist carefully monitor systemic blood pressure as the patient assumes the sitting position. Interventions to maintain blood pressure may include fluid administration, decreasing inhalational agent concentration, compression of distal extremities, and short-term infusions of phenylephrine or other vasoactive drugs. There is a perceived advantage of the sitting position from the point of view of the respiratory system. In this position, access to the chest wall and the airway is obtained, while ventilation is unimpeded because diaphragmatic excursion is greater than in the horizontal position and consequently airway pressure is lower (13). It is important to keep in mind these potential complications when performing skull base procedures in sitting positions. A vital role for the anesthesiologist who is taking care of the sitting patient is to assure that the entire care team participates in the careful positioning of the patient. Goals are to avoid large changes in blood pressure, impairment of ventilation, and avoidance of potential nerve compression and entrapment. Recent reviews have gone as far as to list contraindications for assuming this position, including advanced age, hypertonia, chronic obstructive lung disease, and diagnosed patent foramen ovale (13).
Table 1 Four Grades of Intensity of Venous Air Embolus (VAE) Grade I: characteristic changes in Doppler sounds Grade II: changes in the Doppler sound plus fall of end-expiratory CO2 concentration by more than 0.4% Grade III: changes in Doppler sounds, fall in end-expiratory CO2 concentration, plus aspiration of air through the atrial catheter Grade IV: combination of above signs with arterial hypotension over 20% and/or arrhythmia or other pathological ECG changes Source: From Ref. 14
COMPLICATIONS DURING SKULL BASE PROCEDURES Venous Air Embolism VAE is one of the complications that concern most physicians during skull base surgery. Since skull base procedures commonly deal with venous structures that do not collapse and typically are held open by a bone, it is easy to see that venous air can be entrained via one or several of these open venous plexuses. The occurrence of VAE is especially concerning in patients with patent foramen ovale (present in 10–30% of the population), in which the potential for paradoxical venous air embolus is high. The concern is that changes in hemodynamics may result in right atrial pressures rising higher than left atrial pressures, leading to significant systemic complications such as cerebral infarction. Matjasko and colleagues have described four grades of VAE seen during sitting craniotomies. These grades (I–IV) relate to changes in precordial Doppler sounds, changes in end-tidal carbon dioxide (ETCO2 ) levels, the ability to aspirate air, and the presence of hemodynamic instability (Table 1) (14). Factors that play a role in the development of VAE include the pressure gradient between the surgical field and the right atrium, the surgical technique, and the amount of air entrained. In addition, many neurosurgical patients are hypovolemic, which reduces central venous pressure and further increases the risk of a VAE. Although frequently detected [there was a 72% incidence of VAE detected by transesophageal echocardiography (TEE) in one study (15)], the overall morbidity of VAE is currently low at less than 0.36%. Monitoring for VAE has been extensively described and usually consists of one or more of the following devices (in descending order of sensitivity): TEE, precordial Doppler (in the right third to sixth intercostal space), pulmonary artery pressure, ETCO2 /end-tidal nitrogen, right atrial pressure, electrocardiography (ECG), and esophageal stethoscope (13). Although many clinicians have used a transesophageal echo for detection of VAE because of increased sensitivity, this increase in sensitivity produces too many false positives reducing the specificity as a monitor to detect air entrainment (16). In addition, TEE is expensive and requires trained personnel to be available for interpretation. TEE has also been associated with vocal cord paralysis from recurrent laryngeal nerve palsy after prolonged use. Recommended therapeutic measures in the event of air entrainment include bilateral compression of the jugular veins, having the surgeon flood the wound with saline, discontinuation of nitrous oxide, aspiration of air from a central catheter, and downward tilt of the surgical table (17). The purpose of the atrial catheter is to allow efficient aspiration of trapped air. As such, it is essential that the right atrial catheter be multiorificed in nature. The catheters can be difficult to place, since it is necessary that the catheter tip be in the right atrium, a position that is hard to confirm unless one uses radiographic imaging or ECG tracing (biphasic P-wave morphology). A recent prospective trial regarding the placement
Chapter 6: Patients with Tumors of the Skull Base
of right atrial air aspiration catheters noted that the intravenous ECG P-wave morphology that correlated with the right atrial superior vena cava junction, identified by TEE, was the largest monophasic negative P wave without any biphasic component (18). We prefer the use of an angiographic catheter (Swan-Ganz Angiographic Catheter, Baxter Healthcare Corp, Irvine, CA). This multiorifice catheter is placed in a similar fashion to that of a Swan-Ganz and requires the use of a small pilot balloon at the tip to float the catheter into the right ventricle. Once an right ventricular (RV) waveform is observed, the catheter is withdrawn 1 to 2 cm until the trace disappears. This places the tip in the proper position in the right atrium. Another common practice is changing the position of the patient to left lateral recumbent; however, this has been refuted by Geissler and colleagues (19), who noted that body position had no effect on hemodynamics, nor did the occurrence of right heart failure appear to be related to outflow obstruction. Hypotension and a decrease in coronary perfusion pressure appeared to play more of a role in explaining the cardiovascular effects of air emboli. The role of positive end-expiratory pressure (PEEP) has also come into question. Schmitt and colleagues (15) noted that when PEEP was released, there was a significant occurrence of air emboli as documented by TEE. The group postulates that a sudden decrease of moderate PEEP might decrease right atrial pressure and subsequently increase venous return from cerebral veins. This would result in an increase in air entrainment and possibly an increase in the detectable number of VAE.
Pneumocephalus Another complication associated with skull base surgery is pneumocephalus, defined as the presence of intracranial air. In a retrospective review that compared sitting, park-bench, and prone positioning, 100% of the patients in sitting, 73% of the patients in park-bench, and 57% of the patients in prone positioning had evidence of intraventricular air (20). This has been attributed to the large amount of cerebrospinal fluid (CSF) drained due to gravitational effect. A patient in the sitting position is subject to the effects of gravity more than other positions. Ultimately, more CSF is drained, leading to the high incidence of pneumocephalus seen in the sitting patient. Not every case of intracranial air results in tension pneumocephalus. Historically, approximately 3% of sitting posterior fossa cases are noted to have developed tension pneumocephalus (8). Postoperative care and length of surgery play a role in the development of tension pneumocephalus; however, single contributing factors such as preexisting ventriculo-peritoneal shunts, the utilization of nitrous oxide, or intraoperative diuretics do not appear to play a solitary role (20).
Macroglossia/Facial Swelling A rare, but potentially catastrophic complication of skull base surgeries includes the development of macroglossia. Several case reports describe the occurrence of macroglossia in the immediate postoperative period, but the overall incidence is around 1% (21). Venous drainage of the face, tongue, larynx, and orbits enters the internal jugular system, which, when the neck is maximally flexed, may kink and lead to partial or complete obstruction of the system. In the worse case scenario, this may lead to thrombosis of the internal jugular vein (22). Other theories regarding the etiology of macroglossia also exist, including arterial compression, a neurogenic event,
121
reperfusion injury, and impaired lymphatic drainage (5,23). Lam and colleagues (5) believe that there is a significant role for reperfusion injury since many of the cases are not associated with cerebral swelling and edema, a condition one would expect if obstruction of venous drainage of the internal jugular system was present. The risk factors of obesity, neck flexion, local compression, and long surgical duration have been identified and should be kept in mind when patients present for skull base procedures. Regardless of the specific etiology of macroglossia, careful positioning of the head and neck is essential. As a rule of thumb, the authors assure a space of approximately 2 fingerbreadths between the mandible and the clavicle to prevent venous occlusion.
Cerebrovascular Complications Fortunately, cerebrovascular accidents and complications are rare during skull base procedures. Injury to the carotid artery is one of the most feared complications and may lead to stroke and other brain injuries. Repair of the carotid may be required and may include: saphenous vein bypass graft from the extracranial carotid artery to the petrous carotid artery and superficial temporal to middle cerebral artery bypass (26). Usually, preoperatively, when the carotid artery is affected by tumor, a balloon-occlusion test will be performed. This will help to identify those patients who would tolerate the sacrifice of the carotid artery if it became necessary during the procedure. “Blowout injury” is another carotid vascular complication that may occur. This situation is caused when the carotid artery is inadvertently lacerated as a bone spur is being removed. Blood loss may be brisk at the time of the injury and interventions include packing and urgent angiography with balloon occlusion (27). Delayed blowout injury can also occur if the carotid artery is exposed in the nasopharynx; however, this potentially can be prevented with muscle flap coverage (26). Vasospasm has also been reported and may result in stroke. Vasospasm tends to be seen in younger patients and felt to be due to a myogenic reaction in the vessel wall. It may result from arterial contact with fresh blood or arterial traction. Treatment usually consists of topical vasodilators, such as papaverine, though systemic drugs may have a role (28).
Arrhythmias Since many skull base procedures are in the area of the trigeminal and vagus nerves, as well as the brain stem, arrhythmias during the surgical procedure are common. Direct stimulation of the vagus may lead to negative chronotropy and inotropy manifested as sinus bradycardia, bradycardia terminating in asystole, asystole with no bradycardia, and arterial hypotension. When the trigeminal nerve is involved, sensory nerve endings of the trigeminal send signals to the sensory nuclei at the Gasserian ganglion. These signals ultimately continue along the short internuncial nerve fibers to connect with the efferent pathway in the motor nuclei of the vagus nerve. Again, the effects of stimulation of the cardioinhibitory efferent fibers of the vagus are seen (29). In this situation, the anesthesiologist should alert the surgeon and may request that the surgeon release traction or may choose pharmacologic intervention with a vagolytic substance, such as glycopyrrolate or atropine. Atropine, due to its quick onset, may be the drug of choice; however, the duration of action is shorter than that of glycopyrrolate. Over time, the reflex tends to decrease in its intensity. If the skull base surgery involves resection of a glomus jugulare tumor, one needs to be aware if the tumor is
122
Jellish and Edelstein
catecholamine secreting. Glomus jugulare tumors may be considered arteriovenous malformations, may be giant in size, and may be associated with multiple paragangliomas or adrenal tumors. The incidence of catecholamine secretion is around 4% and can be associated with tachyarrhythmias, hypertension, sweating, myocardial infarction, and cardiovascular collapse, which reflects catecholamine excess. It is important for the anesthesiologist to be prepared to treat any hypertensive crises arising from manipulation of a catecholamine secreting glomus jugulare tumor. Appropriate invasive arterial monitoring is essential and utilization of fast- and short-acting vasodilators, such as nitroglycerin, nitroprusside, and to a lesser extent phentolamine. As seen in the surgical cases involving resection of catecholaminesecreting pheochromocytomas, significant hypotension may occur once the output of norepinephrine has been interrupted. As such, short-term utilization of direct arterial vasopressors (such as phenylephrine or norepinephrine) may be needed to maintain adequate cerebral and coronary perfusion pressures. The resection of glomus jugulare tumors may also result in new cranial nerve injuries of up to 7.1%, with the highest being cranial nerve VII and IX (30).
Blood Loss Blood loss during resection of skull base tumors may be significant and the appropriate preoperative measures for such an event are essential. Many of the tumors are highly vascular in nature and preoperative assessment is necessary to delineate involvement of the cavernous sinus and jugular bulb. Of particular note, meningiomas have been shown to produce tissue plasminogen activator and may lead to increased fibrinolysis during resection (31). Other skull base lesions, known for their vascularity, include jugular paragangliomas, vagal paragangliomas, and nasopharyngeal angiofibromas. Preoperative embolization of feeding vessels may lead to decrease blood loss via a decrease in blood flow and pressure through the tumor (26). Maintenance of normovolemia is essential in skull base operations. When to transfuse will depend on the patient’s coexisting disease processes and is always indicated when there are signs of inadequate oxygen delivery to the tissues. Numerous blood conservation techniques are available including utilization of cell salvage devices in the presence of benign tumors, induced hypotension, acute normovolemic hemodilution, and antifibrinolytic therapy. Risks and benefits of each of the conservation techniques must be assessed prior to institution of therapy.
Peripheral Nerve/Cranial Nerve Injuries Protection from peripheral nerve injuries is of major concern during skull base procedures. The peripheral nerves are at particular risk given the multitude of positions assumed during surgery. All positions from supine to seated have been at one time or another associated with nerve injuries. Varying degrees of injury have been described and the Seddon Classification describes three broad classifications of nerve injury. These include neurapraxia, axonotmesis, and neurotmesis. Neurapraxia is a mild insult that results in conduction failure across an affected segment. This is reversible and tends to be the type of injury most seen during surgical procedures. Axonotmesis is when the axon is physically disrupted but the epineurium and perineurium are preserved. Recovery depends on the speed of neural regeneration. The worst injuries are those in which neurotmesis has taken place. In this situation, there is a complete disruption of the nerve and sup-
port structures and the prognosis for recovery is exceedingly poor (32). Although brachial plexus injuries have been described and felt to be secondary to cervical contralateral flexion rotation and lateral rotation of the shoulder and fixation of the shoulder girdle in a neutral position (33), other injuries have been noted including common peroneal nerve injuries leading to footdrop (34). Ulnar nerve injuries may also occur, but the complete etiology of their occurrence has yet to be determined. The role of extension, rotation, obesity, preexisting disease states, such as diabetes mellitus, have been mentioned as possible contributing factors for ulnar nerve injury. Whatever the etiology, meticulous documentation of appropriate intraoperative padding and the awareness of the existence of preexisting neurological defects is essential when taking care of these patients. Peripheral nerves are not the only nerves at risk during skull base procedures. It is well known that cranial nerve injuries are possible, especially when the operation is close to nerve origin. As mentioned earlier, anosmia is a complication of anterior craniofacial resections and may also lead to changes in taste. Transphenoidal approaches for tumor resection may lead to ocular and oculomotor nerve injuries. This may result in the loss of vision or the development of diplopia. Another major cranial nerve injury encountered during skull base procedures involves the vagus. Vagal injuries result in problems with dysphagia and the potential for aspiration. High vagal nerve paralysis can manifest as true vocal cord paralysis, palate paralysis, loss of pharyngeal muscle function, loss of pharyngeal sensation, and failure of cricopharyngeus to relax (26). Postoperative swallowing evaluation and laryngoscopy is indicated to assess these patients prior to allowing oral intake.
MONITORING AND ANESTHESIA Monitoring for skull base procedures depends on the type of procedure to be performed, the vascular and nerve structures that are involved, and the position in which the patient will be placed for the surgery. In all instances, the patient will have standard routine monitoring, such as ECG, noninvasive blood pressure, pulse oximetry, capnography, and temperature. Other monitors are added as the complexity, blood loss, surgical trauma, and comorbidities of the patient are also factored. Monitoring for skull base procedures must assure for adequate central nervous system perfusion, maintenance of cardiovascular stability and the integrity of the neurologic pathways that are being manipulated. Arterial line placement is the standard for most intracranial and extracranial procedures of the skull base. Invasive blood pressure monitoring provides for closer control of blood pressure and better titration of hyperventilation and blood pH. In addition, hemoglobin levels and electrolyte abnormalities can be easily detected by following serial arterial blood samples obtained from this catheter. Central access with either large-bore single lumen or double lumen catheters is contingent on the length of the surgery, anticipated blood loss, need for estimation of central vascular volume, and position of the patient. Depending on tumor type and location, neurophysiologic monitoring may also be employed to detect disruption of neural tracts or trauma to cranial nerves that may be near the site of surgery. Cranial nerve monitoring has markedly decreased postoperative morbidity after skull base surgery. Electromyography (EMG) of the facial nerve, vagus
Chapter 6: Patients with Tumors of the Skull Base
or trigeminal nerve is used during surgical resection to identify the nerve and preserve neurologic integrity, especially if the nerve is surrounded by a tumor (35). EMG of cranial nerves provides early recognition of surgical trauma, facilitates tumor excision, identifies nerve dysfunction, and confirms nerve function after the tumor is removed. The use of muscle relaxants in conjunction with facial nerve monitoring may be problematic. Muscle relaxants prevent movement and reduce the amount of anesthetic agents needed. At any given level of neuromuscular blockade, a facial muscle response is more resistant to neuromuscular blocking agents than a peripheral muscle (36). This is due to larger motor unit size and the increased number of neuromuscular junctions in the facial muscles. Several studies have demonstrated that neuromuscular blockade, titrated to a T1 of 25%, still allows adequate response from compound motor action potentials of facial muscles to adequately monitor nerve function (37). Nerve irritation and tumor infiltration, however, can lead to reduced or blocked conduction. Preoperative partial facial paralysis may make the monitoring difficult. In addition, external or mechanical noise artifacts could mask muscular contraction and if high-dose inhalational agents are used, muscle activity with nerve stimulation can be further reduced. If muscle relaxants must be used to facilitate surgical resection, alternative methods to monitor the facial nerve may be used. The first method stimulates the nerve at the stylomastoid foramen (38). Antidromic responses are recorded in the operative field. However, this method is awkward to use and does not provide the information obtained by continuous recording. Nerve action potentials can also be recorded at the stylomastoid junction; however, this technique is an evoked response. There is no audible feedback compared to EMG and no information concerning this method’s sensitivity in detecting injury. The final method that may be used to determine facial nerve integrity in the presence of muscle relaxants is the brain stem facial evoked response (39); this nerve monitoring method is based on cross auricular responses to sound that controls ear movement. The facial nerve response is recorded at the mastoid after sound stimulation of the contralateral ear. This technique is technically challenging because of the need for digital computer filtering. The best conditions for monitoring cranial nerve function would be with an anesthetic technique that avoids muscle relaxants. Newer inhalational agents such as sevoflurane and desflurane have low blood gas partition coefficients and reduced fat solubility. These agents produce a rapid induction and emergence with minimal accumulation of anesthetics even after prolonged anesthesia. In addition, short-acting opioids such as remifentanil are extremely potent and have minimal accumulation when given as an infusion. The half life is three to eight minutes and metabolism is produced by nonspecific esterases. The anesthetic technique that produces the best operating conditions for surgery and cranial nerve monitoring uses an anesthetic induction with propofol and fentanyl coupled with an intermediate acting muscle relaxant to facilitate intubation. Maintenance of anesthesia is provided with low-dose desflurane or sevoflurane administered in a 50:50 air/O2 mixture with a background infusion of either fentanyl 2 µg/kg/hr or remifentanil 0.25–0.35 µg/kg/min. No further muscle relaxants are administered as the initial dose is reversed by the time surgery begins. Brain stem auditory evoked potentials (BAEPs) and somatosensory evoked potential (SSEP) monitoring may also be used during resection of skull base tumors, especially if there is possible vascular compromise or ischemia due to temporary occlusion of vascular structures or manipulation
123
of the brain stem. Recording short-latency auditory responses by electrocochleography determines the status of the cochlea and nerves peripheral to the tumor while BAEPs monitor activity central to the tumor (40). Headphone stimulators are unacceptable in the operating room because of their size. Small electrodes are used, the most important of which is placed near the mastoid, on or in the ear. The reference electrode is placed at the top of the head and the ground electrode is placed on the forehead. In general, anesthetic affects on the brain stem auditory evoked response are not dramatic. Slow shifts may be seen as the concentration of inhalational agents increase. Since these recordings are of small amplitude, thousands of responses must be recorded to acquire an adequate average. Frequently, the responses are abnormal and smaller than normal due to the effects of the tumor. Also the time interval required to acquire sufficient responses may reduce the sensitivity of this technique in determining neural injury during tumor removal. Of primary importance for intraoperative monitoring are waves I, III, and V. The interpeak latency I–III provides information regarding the integrity of the peripheral component of the auditory pathway. Shortlatency BAEPs are usually resistant to both intravenous and inhalational agents. Increasing blood levels of barbiturates and ketamine will increase interpeak latency. Propofol however, given at 2 mg/kg/bolus followed by an infusion will increase the latency of I, III, and V waves by <5% without changes in amplitude (41). Inhalational agents such as isoflurane, sevoflurane, and desflurane also increase the latency of the waveform without an appreciable change in amplitude. Hypercapnia does not change BAEP’s waveforms but hypoxia in the presence of normal perfusion will depress waveform amplitude (42). If a decrease in perfusion pressure accompanies hypoxia, all evoked responses become depressed or isoelectric. Body temperature will also affect BAEP with reduced temperature prolonging latency of the observed waveform. Brain stem auditory evoked response amplitude is variably affected by hypothermia. In some studies, amplitude increased with progressive hypothermia reaching a maximum at 28–38 ◦ C (43). At a temperature below 23 ◦ C, however, BAEP waveforms disappeared. In most instances where changes in waveform were due to anesthetics, the observations will occur bilaterally. Unilateral decreases in waveform activity would be indicative of nerve damage secondary to surgical manipulation. The major advantage of intraoperative monitoring with electrocochleography is the detectability of the response. Electrocochleography has a much larger amplitude and these responses can be detected clearly within a few seconds, while detection of BAEP requires minutes. The two recording techniques compliment each other in terms of the regions of the auditory system monitored. There are a number of technical problems that must be overcome with electrocochleographic monitoring. It is necessary to position a recording electrode against the promontorium of the middle ear. As such, this electrode could damage the eardrum or become easily dislodged. The N1 wave is exceedingly large and its loss inevitably predicts postoperative deafness. Both electrocochleography and Brainstem Auditory Evoked Responses (BAER) are considered short-latency responses and, as such, are not markedly affected by inhalational anesthetics. Anesthetic techniques utilizing 1 minimum alveolar concentration (MAC) or less of inhalational anesthetic with or without the addition of nitrous oxide (N2 O) combined with an opioid infusion provide optimum conditions for monitoring these responses to assess cochlear activity.
124
Jellish and Edelstein Table 2 Effect of Inhaled Anesthetics on Somatosensory Evoked Potentials Early cortical waveform Anesthetic drug/concentration Halothane (24, 26, 34) 0.5 MAC + 60% N2 O 1.0 MAC + 60% N2 O 1.5 MAC + 60% N2 O 1.5 MAC (alone) Isoflurane (23–28,31,35,36) 0.5 MAC + 60% N2 O 0.5 MAC (alone) 1.0 MAC + 60% N2 O 1.0 MAC (alone) 1.5 MAC + 60% N2 O∗ 1.6 MAC (alone)∗ Enflurane (24–26) 0.5 MAC + 60% N2 O 0.2–0.6 MAC (alone) 1.0 MAC + 60% N2 O∗ 1.5 MAC + 60% N2 O 1.5 MAC (alone)∗ Sevoflurane (32,33) 0.5 MAC + 66% N2 O 1.0 MAC + 66% N2 O 1.5 MAC + 66% N2 O 1.7–2.5 MAC Desflurane (38,39) 0.5 MAC 1.0 MAC 1.5 MAC Any with 65% N2 O† Nitrous oxide (39,41,47) 60–65%
Latency
Amplitude
Subcortical waveform
<10% ↑ <10% ↑ 10–15% ↑ 10–15% ↑
≈60% ↓ ≈70% ↓ ≈80% ↓ ≈70% ↓
Negligible Negligible Negligible Negligible
<10% ↑ <15% ↑ 10–15% ↑ 15% ↑ >15% ↑ 15–20% ↑
50–70% ↓ <30% ↑ 50–75% ↓ ≈50% ↓ >75% ↓ 60–70% ↓
Negligible Negligible Negligible Negligible 5% ↑ in latency 5% ↑ in latency 20% ↓ in amplitude
<10% ↑ <10% ↑ 20% ↑ Not recordable >25% ↑
≈50% ↓ <20% ↓ ≈85% ↓ Not recordable ≈85% ↓
Negligible NA Negligible Negligible Negligible
<5% ↑ <10% ↑ <10% ↑ 10–15% ↑
38% ↓ ≈45% ↓ ≈50% ↓ ≈100% ↑
Negligible Negligible Negligible NA
<5% ↑ 3–8% ↑ ≤10% ↑ ≥15% ↑
<20% ↓ 30–40% ↓ <50% ↓ >60% ↓
Negligible Negligible Negligible Negligible
No effect
50–55% ↓
Negligible
NA = data not available; negligible = less than 5% change in latency; ↑ = increase; ↓ = decrease. All data are from humans. ∗ In a substantial fraction of patients, waveforms were not attainable at this concentration. † Complete loss of waveform observed only with 1.5 minimum alveolar concentration (MAC) desflurane plus 65% nitrous oxide (N2 O). Source: From Ref. 45.
Finally, somatosensory evoked responses may also be used to assess the integrity of the brain stem and other subcortical structures during surgical procedures involving the skull base. SSEP represents reproducible electrical activity of cortical and subcortical structures locked to a peripheral nerve stimulus. Diagnostic criteria to evaluate intraoperative waveform changes, diagnostic of spinal cord or brain stem dysfunction, are difficult to establish. A decrease in amplitude of 50% or greater and an increase in latency of 10% or greater constitute a significant change that should be investigated (44). Because SSEP has a central component to its measurement, it is very sensitive to the effects of anesthetic agents. All volatile anesthetics produce a dose-dependent increase in SSEP latency, an increase in central conduction time, and a decrease in amplitude (Table 2) (45). Satisfactory monitoring of early cortical SSEPs is possible with 0.5 to 1.0 MAC of halothane, enthrane, and isoflurane. With deep inhalational anesthesia (1.6 MAC), the early cortical N-20 waveform was reproducible but the amplitude was severely decreased (46). The effect of volatile agents on cortical SSEP amplitude is compounded by N2 O. In the presence of 1 MAC isoflurane and N2 O, cortical SSEP demonstrated a 75% decrease in waveform amplitude (47). The newer volatile anesthetics, desflurane and sevoflurane, affect SSEPs like isoflurane but
permit the use of higher inhaled concentrations (upwards of 1.5 MAC). How volatile anesthetics differ quantitatively is still not totally known, but sevoflurane and desflurane are associated with less amplitude reduction than isoflurane at a MAC range of 0.7 to 1.3 (48). Intravenous anesthetics generally affect SSEPs less than inhaled anesthetics. SSEP waveforms are preserved even at high doses of narcotics and barbiturates. Intravenous agents only modestly affect early and intermediate SSEP components. Most authors report clinically unimportant changes in SSEP latency and amplitude with the administration of opioids given in either anesthetic or analgesic doses. McPherson et al. found minimal SSEP changes after 25 µg/kg fentanyl (49). They noted a small increase in cortical median nerve SSEP latency and a variable decrease (0–30%) in amplitude after 36 to 71 µg/kg fentanyl was administered. This makes opioids useful as part of the anesthetic technique when intraoperative neurophysiologic monitoring is used. Propofol affects the SSEP waveform by increasing latency and decreasing central amplitude in a dose-dependent manner. This drug affects synaptic transmission more than axonal conduction. However, when given as either a bolus or an infusion, it can still be used to monitor central SSEP. When used as a sedative hypnotic along with opioids, propofol reduces SSEP amplitudes less than N2 O or
Chapter 6: Patients with Tumors of the Skull Base
midazolam. Anesthetics that result in latency prolongation or amplitude depression may confuse the interpretation of SSEP changes and potentially risk either not detecting a critical event or providing excessive false positive interpretations. These anesthetic regimens should be avoided. Anesthetic techniques where high-dose inhalational agents, at greater than 1 to 1.3 MAC, in combination with N2 O are problematic. The substitution of propofol for N2 O increases cortical SSEP amplitude up to 100% during an opioid based anesthetic. Substituting remifentanil for fentanyl and N2 O, during a low-dose isoflurane anesthetic, decreases SSEP waveform variability, which should improve reliability. The effect of anesthetics on evoked potentials can be greater in neurologically impaired patients than in those without preoperative deficits. Thus the anesthetic regimen must be adjusted to carefully limit the concentration of volatile anesthetics to less than 1 MAC or avoid N2 O. If neurophysiologic monitoring will include electrocochleography or SSEP, volatile anesthetics with N2 O should be limited to 0.5 MAC concentrations. Without N2 O, 1 MAC concentrations may be used. Newer volatile anesthetics such as desflurane or sevoflurane seem to allow neurophysiologic monitoring at higher concentration but may produce baseline nerve firing with facial nerve monitoring. The use of continuous infusion of intravenous anesthetics (propofol) and opioids with low concentration of background inhalational anesthetics is ideal and recommended for intraoperative neurophysiologic monitoring during skull base procedures.
ANESTHETIC CONCERNS WITH VASCULAR LESIONS Skull base surgeries that have a vascular component may present many problems concerning intraoperative anesthetic management. Complete resection to alleviate symptoms or affect cure while minimizing neurologic complications may be difficult. Vascular bypass procedures may be done for many pathologic entities including aneurysms that encompass the internal carotid or tumors that involve the artery at the skull base or cervical region. Aneurysms may form in the petrous cavernous portion of the skull base where repairs and access to the lesion are difficult. In other situations, paragangliomas may also involve the carotid artery, both in the cervical and in skull base region. If the carotid artery is to be occluded during the surgery, some form of assessing cerebral perfusion may be necessary. In addition, if the possibility of cerebral hypoperfusion exists, methods to provide cerebral protection must be incorporated into the anesthetic plan. As mentioned earlier, cerebrovascular complications during and after these procedures may occur. After resection of tumor, involving the carotid, complications are thought to occur by two mechanisms: the first is by acute infarction after interruption of blood flow; the second is infarction as a result of clot propagation or embolic phenomenon from static blood. Despite advances in surgical techniques and preoperative assessment, patients are still at risk for transient or permanent stroke, coma, or death from thromboembolic phenomenon or sustained hypoperfusion. For these reasons, use of perioperative anticoagulation and antiplatelet therapy, plus careful intraoperative blood pressure control with brain protection measures is important. If the skull base procedure is likely to cause interruption of cerebral blood flow, the intraoperative anesthetic management can be tailored to produce optimal perfusion to the areas affected by the interruption of blood flow. The patient should be hemodiluted prior to interruption of perfusion to reduce blood viscosity and to increase
125
blood flow (50). The optimal hematocrit should approximate 30% to provide the best oxygen transport and cerebral oxygen metabolism. Glucose concentrations are also important if there is a risk of cerebral ischemia. High plasma glucose levels induce anaerobic glycolysis and lactic acidosis, which results in brain edema and microcirculatory dysregulation (51). In addition, insulin protects against ischemicinitiated excitatory damage, protects mitochondrial oxidation capacity, decreases serum K+ levels and glucose utilization, and directly stimulates Na+ /K+ -ATPase (52). Tight control of plasma glucose is warranted for perioperative outcome. Plasma glucose should be maintained between 95 mg/dl and 105 mg/dl throughout the intraoperative period.
Anesthesia and Neuroprotection If temporary vascular occlusion is anticipated, the anesthetic technique should incorporate strategies to augment systemic blood pressure. Cerebral ischemia impairs autoregulation of the cerebral vasculature resulting in resistance blood vessels that are maximally dilated. Blood flow to the ischemic region becomes pressure dependent. Increasing perfusion pressure improves cerebral blood flow and may reduce cell death in the compromised vascular territory by improving collateral circulation (53). Phenylephrine, norepinephrine, and dopamine are the most commonly used agents. Phenylephrine does not cause direct cerebral vasoconstriction and is often the drug of choice. The ability of phenylephrine to augment cerebral blood flow is probably due to a systemic effect. The tendency for a reflex decrease in heart rate makes the use of phenylephrine, particularly, useful in patients with coronary artery disease. Another simple yet effective measure to reduce cerebral ischemia during temporary disruption of blood flow is to elevate the inspired fraction of oxygen. This results in higher tissue levels than expected on dissolved oxygen in blood. Increasing numbers of studies support the use of normobaric 100% O2 in the setting of transient focal ischemia (54). The neuroprotective effects of hypothermia are not precisely known, but probably relate to reduced cerebral oxygen demand (7–8% per 1 ◦ C), decreased release of excitatory neurotransmitters, and increased release of inhibitory neurotransmitters (55). Evidence indicates that the relationship between brain protection and the degree of hypothermia is not linear and mild hypothermia (34 ◦ C) may provide protection against cerebral ischemia (56). Deliberate hypothermia to 34 ◦ C is easily achieved intraoperatively by passive cooling and can be reversed after the restoration of blood flow by conventional methods (convection heaters, circulatory water blankets, forced-air warners, etc.). Besides the physiologic manipulations that can be used to preserve the brain during temporary interruption of blood flow, pharmacologic agents may also be used to decrease neuronal activity and metabolic demand. The use of thiopental to protect the brain has been demonstrated during cardiac bypass (57). Thiopental, at doses capable of producing electroencephalogram (EEG) burst suppression, should offer protection during short occlusion times and longer with the use of continuous infusions of the drug. Barbiturates are also thought to enhance gamma-aminobutyric acid activity and antagonize the N-methyl-D-aspartate receptor which reduces ischemic excitotoxicity (58). Propofol, like barbiturates, will induce burst suppression in a dose-dependent fashion. Furthermore, it is metabolized quickly and therefore does not accumulate. Although propofol has been shown to be of some benefit, results of animal studies have been inconclusive compared with thiopental (59). Etomidate has also been shown to prevent increases in excitatory neurotransmitters
126
Jellish and Edelstein
during cerebral ischemia; however, its protective role compared to other anesthetics is unclear. Recent studies have demonstrated that etomidate administration, prior to cerebral ischemia, produces a 30% reduction in PaO2 levels and a 23% increase in PaCO2 concentrations in the cerebral cortex (60). These changes were thought to be due to etomidate associated vasoconstriction and a reduction in cerebral blood flow prior to a drop in cerebral metabolic rate. This reduces etomidate’s effectiveness as a neuroprotectant, even though its short half life and hemodynamic stability is superior to barbiturates. Volatile anesthetics can also be used to reduce cerebral metabolism, but their ability to dilate arteries and reduce blood pressure makes them a less than optimal choice when attempting to maintain cerebral perfusion to ischemic areas. Other agents such as calcium channel blocker and local anesthetics may provide some added benefit through their ability to block ion channels and reduce Na+ /K+ transmembrane flux which reduces basal energy expenditures (61). Most effective neuroprotective techniques used today involve some combination of pharmacologic therapy coupled with some level of hypothermia or other physiologic manipulation. If neuroprotection is needed or anticipated, EEG monitoring may be helpful for detection of cerebral ischemia or for titration of anesthetic agents used to produce burst suppression. In addition to EEG, SSEP monitoring of cerebral function has been advocated (62). Although there is a clearly documented relationship between electrophysiologic changes and decreased blood flow, scalp electrodes placed at a distance to the ischemic area are relatively insensitive. The electrical signal is further attenuated by CSF drainage and osmotherapy, which introduces air between the scalp electrodes and the underlying cerebral cortex. SSEP monitoring for cerebral ischemia during temporary vessel occlusion is limited by its inability to detect ischemia in the motor cortex, subcortical structures, and sensory regions not topically represented by the stimulated peripheral nerve. In many instances, brain stem auditory evoked potential monitoring may also be beneficial, especially if the posterior circulation is involved. Whatever the monitoring technique used, the anesthetic method must be adjusted to provide the best conditions for monitoring.
ANESTHESIA FOR ENDOSCOPIC SKULL BASE SURGERY As in most fields of medicine, there has been a movement toward less invasive procedures. As new imaging devices and surgical techniques have developed, there has been an explosive growth in the area of minimally invasive neurosurgical techniques. One area that has seen this development is endoscopic skull base surgery, especially with acoustic neuromas and pituitary tumors. The traditional translabyrinthine approach to acoustic neuromas results in the sacrifice of hearing during the course of the procedure and other complications, including stroke, hemorrhage, neuropathies, CSF leaks, and cranial nerve injuries. Much as the lateral approach to the skull improved the ability to access certain skull base tumors and reduce postoperative morbidity, endoscopic skull base surgery has become a popular technique to access skull base pathology and resect the lesion with minimal operative trauma and an improved postoperative course. The endoscopic approach requires an image guidance system that may either be computer tomography or magnetic resonance based. The imaging system helps to map the
area of interest and guide in the resection of the tumor in an efficient manner. In addition, specific instruments have been created to maximize efficiency in a tight space (63). The patient is usually in a semi-sitting position and rotated away from the anesthesia team, thus the airway is removed from the immediate proximity of the anesthesiologist. Endoscopic endonasal approaches to the anterior, clival, and posterior skull base have reduced the need for postoperative analgesia therapy and the surgical trauma that accompanies these procedures. Endoscopic resections of pituitary, craniopharyngiomas, chordoma, and other tumors have been done in recent years through small incisions through the nostril, glabella, and orbital roof by incisions in the eyebrow, superolateral orbit, and subtemporal regions. Since there is reduced blood loss and surgical trauma, anesthetic techniques have been modified to adjust to this new emerging surgical technology. The general aims of the anesthesiologist are to provide hemodynamic stability, maintain cerebral oxygenation, provide favorable conditions for surgical exposure, and facilitate a rapid emergence. If the pituitary is involved, hormone replacement therapy should be continued into the operative period. In all instances, a careful airway evaluation should be performed, especially if the patient has a pituitary tumor and is secreting abnormal amounts of growth hormone. Acromegalics have particularly problematic airways and depending on the amount of soft tissue overgrowth of the mandible and the amount of tongue enlargement, these patients could be exceedingly problematic to intubate. If awake, fiber-optic intubation cannot be safely performed; a tracheostomy may be done to facilitate control of the airway. With the exception of pituitary tumors, airway control for other endoscopic skull base procedures is routine. The endotracheal tube is secured without tape to the upper lip. For a transnasal endoscopic approach, the surgeon may want to introduce a vasoconstricting agent into each nostril to prevent bleeding. Traditionally, a mixture of cocaine and epinephrine is used. Although the addition of epinephrine limits systemic absorption, the use of cocaine-containing preparations continues to be associated with a risk of arrhythmia and myocardial infarction. We routinely place an arterial line in all patients who will be administered nasal cocaine to monitor blood pressure and heart rhythm during systemic absorption. Most endoscopic approaches to the skull base are done with the patient in the supine head-up position with the head tilted or turned to improve access. The operating room (OR) table could be at 90 degrees from the anesthesia provider or turned to 180 degrees, depending on surgical preferences. In most instances, the trachea is intubated to control respirations and ETCO2 . However, with some endoscopic approaches, a laryngeal mask airway may be used with the patient spontaneously ventilating. Any anesthetic technique is suitable since most of these cases are done without neurophysiologic monitoring. However, if intraoperative neurophysiologic monitoring is to be used, the anesthetic technique must be adjusted to provide optimal conditions for monitoring. The choice of anesthetic technique for these procedures is usually determined by personal preference. A total intravenous technique with propofol and remifentanil provides ideal intraoperative conditions with rapid emergence and extubation. Short-acting inhalational agents can also be used in conjunction with remifentanil. Since postoperative pain is minimal, longer half-life analgesics are not necessary and anesthetic techniques are modified to produce a rapid emergence with minimal hangover effect. With some of the endoscopic approaches, there may be periods of intense stimulation. The short-acting opioids should be titrated against
Chapter 6: Patients with Tumors of the Skull Base
blood pressure. At the end of the endoscopic procedure, the patient should emerge rapidly and have no overt bleeding. If hemodynamically stable, breathing spontaneously, and following commands, the patient may be extubated. Recently, in a study that compared the sublabial transeptal hypophysectomy to the endoscopic approach, there was a significant reduction in the rate of nasal complications and CSF leaks (64). Others have revealed that endoscopic procedures result in less postoperative pain and less blood loss when compared to standard nonendoscopic approaches (65).
POSTOPERATIVE MANAGEMENT After the completion of the skull base procedure, depending on blood loss and surgical length, the patients who have sufficiently recovered from anesthesia to follow commands and maintain oxygenation while spontaneously breathing are usually extubated. Recovery from anesthesia is expedited by the use of short-acting, low solubility inhalational agents and infusions of potent opioids. Cranial nerve deficits, including injury to the vagus nerve producing vocal cord paralysis, may occur after skull base surgery. Even though this would produce unilateral vocal cord paralysis, the patient should be awake and able to handle secretions to avoid the risk of aspiration. Nausea, vomiting, and pain are important aspects for consideration in the immediate postoperative period. The incidence, magnitude, and duration of acute pain experienced after craniotomy are not well known. Pain after acoustic neuroma surgery, with a posterior fossa approach, has been noted to be severe in 67% of patients and was thought to be due to either nuchal dissection or traction on the dura by nuchal musculature (66). Evidence shows that patients who undergo skull base surgeries experience varying degrees of postoperative pain. Some of this variability may be explained by the different anatomical approaches to the brain and meninges. The proximity of cranial nerves and the chemoreceptor trigger zone to the surgical field may increase the risk of pain and nausea after posterior fossa and skull base surgery. Pain from the infratentorial structures is transmitted by efferent fibers in cranial nerve V, IX, and X and the upper three cervical nerves. There is evidence that neurosurgical patients receive inadequate analgesia from currently available regimens (67). The desire to avoid respiratory depression with CO2 retention and preservation of pupillary responses has reduced the use of strong opioid analgesics in this patient population. De Beneditis et al. demonstrated that patients undergoing surgery by the subtemporal and suboccipital routes have the highest incidence of postoperative pain (68). Approximately 90% of these patients experienced pain in the first 12 hours after surgery. The pain that occurred was influenced by surgical route and was defined as pulsating or pounding, steady, and sometimes stabbing in nature. Although no ideal analgesic exists, all believe patient controlled analgesia (PCA) is extremely valuable after skull base procedures (68). PCA has been found to be subjectively better for patients and leads to overall lower doses of opioid. It produces a mechanism to titrate drug administration, allows the patients to control their pain, and may alleviate the psychological stress of pain. Morphine in a dose regimen of 1.5 mg/dose with an eight-minute lockout and a maximum dose of 40 mg at four hours has not been observed to cause respiratory depression requiring reintubation. The most common associated morbidity and reason for dissatisfaction with PCA is the development of nausea and vomiting. This may be due to the type of surgery and
127
anesthetic agents used, but could also be due to blood opioid concentrations that change with PCA administration (69). Although nausea and vomiting may have evolved as protective reflexes, they are undesirable side effects that can lead to significant morbidity such as increased intracranial pressure, loss of fluids/electrolytes, alkalosis, and worsening respiratory function. Physiologic evidence suggests that vomiting may be associated more with infratentorial surgery (67). Since pain is controlled, in most instances, by the judicious use of opioids, subsequent nausea and vomiting is controlledby antiemetics. Several methodologies are employed to provide relief from nausea or vomiting with the antiemetic given as a rescue agent. Droperidol, a dopamingeric antagonist, given in a dose of 0.625 mg IV has been noted to be effective in reducing nausea and the incidence of vomiting (70). However, it may cause extensive extrapyramidal side effects and synergistically increase the sedation associated with opioids. Thus, it is not the ideal antiemetic. Ondansetron, a serotonin receptor antagonist, has also been used with some limited success in reducing the incidence and severity of nausea and vomiting. Finally, the addition of dexamethasone (decadron) to antiemetic regimens seems to reduce the incidence of nausea and vomiting (71). Dexamethasone has been demonstrated to be effective for prophylaxis of postoperative nausea and vomiting secondary to prostaglandin metabolism (72). Others have suggested that the dexamethasone’s effectiveness may be due to a release of endorphins resulting in mood elevation, a sense of well-being, and appetite stimulation. Dexamethasone has also been used in combination with 5HT3 receptor antagonists. Corticosteroids may reduce levels of 5-hydroxytryptophan in neural tissue by decreasing tryptophan, its precursor. Secondly, the anti-inflammatory portion of dexamethasone may prevent release of serotonin in the gut. Lastly, dexamethasone may potentiate the main effects of the antiemetics by sensitizing pharmacological receptors. Antiemetics have also been added to PCA solutions to administer antiemetics simultaneously in conjunction with opioids. Our group demonstrated that patients who undergo craniotomy, especially in the skull base region, have significant postoperative pain (73). The combination of ondansetron with PCA morphine reduced pain scores in the immediate 24 hours post surgery, but had little effect on the incidence of nausea and vomiting, and should not be used in combination with PCA morphine. R Ondansetron ODT oral treatment given prior to surgery with intraoperative IV dexamethasone plus IV ondansetron was associated with less frequent rescue therapy on the first postoperative day compared with patients receiving IV dexamethasone plus placebo (71). This multimodal treatment of nausea and vomiting after skull base procedures may be more effective in reducing nausea and vomiting associated with pain or opioid administration than single therapy regimens. Multimodal analgesic therapies should also be used to reduce pain. Besides aggressive PCA therapy, local anesthetics can be used to infiltrate the surgical wound during closure. Lidocaine 2% with 1:200,000 epinephrine prolongs the analgesic half life while reducing bleeding from the incision site. Other anti-inflammatory agents such as ketorolac or indomethacin may be added to reduce the narcotic demand secondary to musculoskeletal trauma. In conclusion, the anesthesia technique used for skull base surgery is instrumental in providing optimum conditions for intraoperative monitoring while enabling the surgeon to accomplish the surgical procedure in a hemodynamically stable patient. The anesthesiologist can also improve postoperative outcomes and reduce morbidity provided he
128
Jellish and Edelstein
or she is given the information concerning surgical approach, tumor type, and involved vital structures. It is imperative that the neuroanesthesiologist discuss the procedure with the skull base surgical team and be involved in the decisions concerning perioperative management to produce an optimal outcome.
REFERENCES 1. Jellish WS, Murdoch J, Leonetti JP. Perioperative management of complex skull base surgery: The anesthesiologist’s point of view. Neurosurg Focus. 2002;12(5):1–7 2. Jung TM, TerKonda P, Haines SJ, et al. Outcome analysis of the transglabellar/subcranial approach for lesions of the anterior cranial fossa: A comparison with the classic craniotomy approach. Otolaryngology. 1999;116(6):642–646. 3. Smith DS, Osborn I. Posterior fossa: Anesthetic consideration. In: Cottrell JE, Smith D, Cottrell JE, et al., eds. Anesthesia and Neurosurgery. 4th ed. St. Louis: Mosby, 2001: 327–331 4. Faust RJ, Cucchiara RF, Bechtle P. Patient positioning. In: Miller RD, Miller RD, eds. Miller’s Anesthesia. Philadelphia, PA: Elsevier Science Health Science div, 2005: 1151–1167. 5. Lam AM, Vavilala MS. Macroglossia: Compartment syndrome of the tongue? [report]. Anesthesiology. 2000;92(6):1832. 6. Deem S, Shapiro HM, Marshall LF. Quadraplegia in a patient with cervical spondylosis after thoracolumbar surgery in the prone position. Anesthesiology. 1991;75(3):527–528. 7. Black S, Ockert DB, Oliver WC, et al. Tumor surgery. In: Cucchiara RF, Michenfelder JD, Cucchiara RF, et al., eds. Clinical Neuroanesthesia. Edinburgh: Churchill Livingstone, 1990:285–308. 8. Standefer M, Bay JW, Trusso R. The sitting position in neurosurgery: A retrospective analysis of 488 cases. Neurosurgery. 1984;14(6):649–658. 9. Wilder BL. Hypothesis: The etiology of midcervical quadriplegia after operation with the patient in the sitting position. Neurosurgery. 1982;11(4):530–531. 10. Hall R, Murdoch J. Brain protection: Physiological and pharmalogical considerations. Part ii: The pharmacology of brain protection. Can J Anaesth. 1990;37:762–777. 11. Ward RJ, Danziger F, Bobica JJ, et al. Cardiovascular effects of change of posture. Aerosp Med. 1966;37(3):257–259. 12. Millar RA. Neurosurgical anaesthesia in the sitting position. Br J Anaesth. 1972;44(5):495–505. 13. Porter JM, Pidgeon C, Cunningham AJ. The sitting position in neurosurgery: A critical appraisal. Br J Anaesth. 1999;82(1):117– 128. 14. Matjasko J, Petrozza P, Cohen M, et al. Anesthesia and surgery in the seated position: Analysis of 554 cases. Neurosurgery. 1985;17(5):695–702. 15. Schmitt HJ, Hemmerling TM. Venous air emboli occur during release of positive end-expiratory pressure and repositioning after sitting position surgery. Anesth Analg. 2002;94(2):400–403. 16. Guggiari M, Lecha P, Garen-Colonne C, et al. Early detection of patent foramen ovale by two-dimensional contrast echocardiography for prevention of paradoxical air embolism during sitting position. Anesth Analg. 1988;67(2):192–194. ¨ 17. von Gosseln HH, Samii M, Suhr D, et al. The lounging position for posterior fossa surgery: Anesthesiological considerations regarding air embolism. Childs Nerv Syst. 1991;7:368–374. 18. Kerr RH, Applegate RL. Accurate placement of the right atrial air aspiration catheter: A descriptive study and prospective trial of intravascular electrocardiography. Anesth Analg. 2006;103(2):435–438. 19. Geissler HJ, Allen SJ, Mehlhorn U, et al. Effect of body repositioning after venous air embolism: An echocardiographic study. Anesthesiology. 1997;86(3):710–717. 20. Toung TJK, McPherson RW, Ahn H, et al. Pneumocephalus: Effects of patient position on the incidence and location of aerocele after posterior fossa and upper cervical cord surgery. Anesth Analg. 1986;65(1):65–70.
21. Moore JK, Chaudhri S, Moore AP, et al. Macroglossia and posterior fossa disease. Anaesthesia. 1988;43:382–385. 22. Tattersall MP. Massive swelling of the face and tongue: A complication of posterior cranial fossa surgery in the sitting position. Anaesthesia. 1984;39:1015–1017. 23. Kuhnert SM, Faust RJ, Berge KH, et al. Postoperative macroglossia: Report of a case with rapid resolution after extubation of the trachea. Anesth Analg. 1999;88(1):220–223. 24. Khrais TH, Falcioni M, Taibah A, et al. Cerebrospinal fluid leak prevention after translabyrinthine removal of vestibular schwannoma. Laryngoscope. 2004;114(6):1015–1020. 25. Leonetti J, Anderson D, Marzo S, et al. Cerebrospinal fluid fistula after transtemporal skull base surgery. Otolaryngol Head Neck Surg. 2001;124(5):511–514. 26. Schwaber MK, Netterville JL, Coniglio JU. Complications of skull base surgery. Ear Nose Throat J. 1991;70(9):648–660. 27. Bogdasarian RS, Kwyer TA, Dauser RC, et al. Internal carotid artery blowout as a complication of sphenoid and skull-base surgery. Otolaryngol Head Neck Surg. 1983;91(3):308–312. 28. Smith PG, Killeen TE. Carotid artery vasospasm complicating extensive skull base surgery: Causes, prevention, and management. Otolaryngol Head Neck Surg. 1987;97(1):1–7. 29. Schaller B. Trigemino-cardiac reflex during microvascular trigeminal decompression in cases of trigeminal neuralgia. J Neurosurg Anesthesiol. 2005;17:45–48. 30. Al-Mefty O, Teixeira A. Complex tumors of the glomus jugulare: Criteria, treatment, and outcome. J Neurosurg. 2002;97(6):1356– 1366. 31. Tsuda H, Oka K, Noutsuka Y, et al. Tissue-type plasminogen activator in patients with intracranial meningiomas. Thromb Haemost. 1988;60:508–513. 32. Aminoff MJ. Electrophysiologic testing for the diagnosis of peripheral nerve injuries. Anesthesiology. 2004;100(5):1298–1303. 33. Coppieters MW, Van De Velde M, Stappaerts KH. Positioning in anesthesiology: Toward a better understanding of stretch-induced perioperative neuropathies. Anesthesiology. 2002;97(1):75–81. 34. Sawyer RJ, Richmond MN, Hickey JD, et al. Peripheral nerve injuries associated with anaesthesia. Anaesthesia. 2000;55(10):980– 991. 35. Cheek JC. Posterior fossa intraoperative monitoring. J Clin Neurophysiol. 1993;10:412–424. 36. Lennon RL, Hosking MP, Daube JR, et al. Effect of partial neuromuscular blockade on intraoperative electromyography in patients undergoing resection of acoustic neuromas. Anesth Analg. 1992;75(5):729–733. 37. Blair EA, Teeple E Jr, Sutherland RM, et al. Effect of neuromuscular blockade on facial nerve monitoring. Am J Otol. 1994;15(2):161–167. 38. Stechison MT. Neurophysiologic monitoring during cranial base surgery. J Neurooncol. 1994;20:313. 39. Yingling CD. Intraoperative monitoring of cranial nerves in skull base surgery. In: Jackler RK, Brackmann DE, Jackler RK, et al., eds. Neurotology St. Louis: Mosby, 1994:967–1002. 40. Symon L, Jellinek DA. Monitoring of auditory function in acoustic neuroma surgery. In: Schramm J, Moller AR, Schramm J, Moller AR, eds. Intraoperative Neurophysiologic Monitoring in Neurosurgery. Berlin: Springer-Verlag, 1991:205–213. 41. Chassard D, Colson A, Banssillon V, et al. Auditory evoked potentials during propofol anaesthesia in man. Br J Anaesth. 1989;62(5):522–526. 42. Sohmer H, Gani M, Chisin R. Auditory nerve-brainstem potentials in man and cat under hypoxic and hypercapnic conditions. Electroencephalogr Clin Neurophysiol. 1982;53:506–512. 43. MacKenzie MA, Vingerhoets DM, Colon EJ, et al. Effect of steady hypothermia and normothermia on multimodality evoked potentials in human poikilothermia. Arch Neurol. 1995;52(1):52– 58. 44. McTaggert-Cowan RA. Somatosensory evoked potentials during surgery. Can J Anaesth. 1988;45:387–392. 45. Banoub M, Tetzlaff JE, Schubert A. Pharmacologic and physiologic influences affecting sensory evoked potentials: Implications
Chapter 6: Patients with Tumors of the Skull Base
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
for perioperative monitoring. Anesthesiology. 2003;99(3):716– 737. Porkkala T, Kaukinen S, Hakkinen V, et al. Median nerve somatosensory evoked potentials during isoflurane anaesthesia. Can J Anaesth. 1997;44:963–968. Wolfe DE, Drummond JC. Differential effects of isoflurane/nitrous oxide on posterior tibial somatosensory evoked responses of cortical and subcortical origin. Anesth Analg. 1988;67(9):852–859. Rehberg B, Ruschner R, Fischer M, et al. Concentrationdependent changes in the latency and amplitude of somatosensory-evoked potentials by desflurane, isoflurane and sevoflurane. Anasthesiol Intensivmed Notfallmed Schmerzther. 1988;33:425–429. McPherson RW, Sell B, Thaystman RJ. Effect of thiopental, fentanyl and etomidate on upper extremity somatosensory evoked potentials in humans. Anesthesiology. 1986;65:584– 589. Tu YK, Heros RC, Candia G, et al. Isovolemic hemodilution in experimental focal cerebral ischemia. Part I: Effects on hemodynamics, hemorheology, and intracranial pressure. J Neurosurg. 1988;69:72–81. Wass CT, Lanier WL. Glucose modulation of ischemic brain injury: Review and clinical recommendations. Mayo Clin Proc. 1996;71(8):801–812. Voll CL, Auer RN. Insulin attenuates ischemic brain damage independent of its hypoglycemic effect. J Cereb Blood Flow Metab. 1991;11:1006–1014. Wasnick JD, Conlay LA. Induced hypertension for cerebral aneurysm surgery in a patient with carotid occlusive disease. Anesth Analg. 1990;70(3):331–333. Hoffman WE, Charbel FT, Edelman G. Brain tissue oxygen, carbon dioxide, and ph in neurosurgical patients at risk for ischemia. Anesth Analg. 1996;82(3):582–586. ¨ CH, Smith ML, et al. The influence Minamisawa H, Nordstrom of mild body and brain hypothermia on ischemic brain damage. J Cereb Blood Flow Metab. 1990;10:365–378. Wass CT, Lanier WL. Hypothermia-associated protection from ischemic brain injury: Implication for patient management. Int Anesthesiol Clin. 1996;34:95–111. Nussmeier NA, Arlund C, Slogoff S. Neuropsychiatric complications after cardiopulmonary bypass: Cerebral protection by a barbiturate. Anesthesiology. 1986;64(2):165–170. Michenfelder JD, Milde JH, Sundt TM Jr. Cerebral protection by barbiturate anesthesia. Use after middle cerebral artery occlusion in java monkeys. Arch Neurol. 1976;33:345–350.
129
59. Zhu H, Cottrell JE, Kass IS. The effect of thiopental and propofol on nmda- and ampa-mediated glutamate excitotoxicity. Anesthesiology. 1997;87(4):944–951. 60. Edelman GJ, Hoffman WE, Charbel FT. Cerebral hypoxia after etomidate administration and temporary cerebral artery occlusion. Anesth Analg. 1997;85(4):821–825. 61. Lei B, Cottrell JE, Kass IS. Neuroprotective effect of low-dose lidocaine in a rat model of transient focal cerebral ischemia. Anesthesiology. 2001;95(2):445–451. 62. Lam AM, Manninen PH, Ferguson GG, et al. Monitoring electrophysiologic function during carotid endarterectomy: A comparison of somatosensory evoked potentials and conventional electroencephalogram. Anesthesiology. 1991;75(1):15–21. 63. Manning SC, Bloom DC, Perkins JA, et al. Diagnostic and surgical challenges in the pediatric skull base. Otolaryngol Clin North Am. 2005;38(4):773–794. 64. White DR, Sonnenburg RE, Ewend MG, et al. Safety of minimally invasive pituitary surgery (mips) compared with a traditional approach. Laryngoscope. 2004;114(11):1945–1948. 65. Casler JD, Doolittle AM, Mair EA. Endoscopic surgery of the anterior skull base. Laryngoscope. 2005;115(1):16–24. 66. Schessel DA, Nedzelski JM, Rowed D, et al. Pain after surgery for acoustic neuroma. Otolaryngol Head Neck Surg. 1992;107:424– 429. 67. Irefin SA, Schubert A, Bloomfield EL, et al. The effect of craniotomy location on postoperative pain and nausea. J Anesth. 2003;17(4):227–231. 68. De Benedittis G, Lorenzetti A, Migliore M, et al. Postoperative pain in neurosurgery: A pilot study in brain surgery. Neurosurgery. 1996;38(3):466–470. 69. Watcha MF, White PF. Postoperative nausea and vomiting: Its etiology, treatment, and prevention. Anesthesiology. 1992;77(1):162–184. 70. Henzi I, Sonderegger J, Tramer MR. Efficacy dose-response and adverse effects of droperiodol for prevention of postoperative nausea and vomiting. Can J Anaesth. 2000;47:537–551. 71. Hartsell T, Long D, Kirsch JR. The efficacy of postoperative ondansetron (zofran(r)) orally disintegrating tablets for preventing nausea and vomiting after acoustic neuroma surgery. Anesth Analg. 2005;101(5):1492–1496. 72. Henzi I, Walder B, Tramer MR. Dexamethasone for the prevention of postoperative nausea and vomiting: A quantitative systematic review. Anesth Analg. 2000;90(1):186. 73. Jellish WS, Leonetti JP, Sawicki K. Morphine/ondansetron pca for postoperative pain, nausea, and vomiting after skull base surgery. Otolaryngol Head Neck Surg. 2006;135(2):175– 181.
7 Minimally Invasive Techniques: Endonasal Endoscopic Skull Base Surgery Allan D. Vescan, Ricardo L. Carrau, Carl H. Snyderman, Amin B. Kassam, Arlan Mintz, and Paul Gardner
INTRODUCTION
achieved, either by an “en bloc” resection or by a piecemeal removal, as long as the microscopic analysis of all margins is negative. Our opinion is based on our experience with open skull base approaches in which the “en bloc” removal of tumor is rarely achieved and ultimately the tumor is removed in multiple small pieces. Third, every attempt should be made to preserve major neurovascular structures. As a result of this principle, when approaching tumors via an endoscopic endonasal approach, one of the critical determinants of feasibility and safety of the resection is the relationship of critical neurovascular structures. If a lesion is located in a line of sight that is distal to the internal carotid artery (ICA) or the optic nerve, then one must consider alternate approaches. Fourth, the skull base reconstruction must separate the cranial cavity from the sinonasal tract. In our experience, this has been one of the most challenging and dynamic areas of endonasal approaches to the skull base. We will discuss this issue in detail later in this chapter. The final tenet to be considered when choosing an approach is to be able to minimize morbidity, preserve function and quality of life, and preserve or reconstruct optimal cosmesis (Table 1).
Skull base surgery is a relatively new subspecialty. Since the seminal paper by Ketcham et al. in the early 1960s the anterior craniofacial resection became the standard approach to treat tumors of the anterior cranial base (1). Halstead used the transnasal approach to access the sella turcica for the treatment of pituitary tumors in the early 1900s (2), an approach that was later popularized by Cushing (3). Jules Hardy introduced the microscope as a visualization tool and ultimately standardized the transnasal transseptal approach as the optimal technique to extirpate tumors of the sella (4). With the advent of rod lens telescopes in the late 1970s, numerous pioneering otolaryngologists, including Messerklinger, Stammberger, Wigand, Kennedy, and others, successfully applied endoscopic techniques to treat disorders of the paranasal sinuses (5–8). The logical progression, as otolaryngologists became familiar with the use of endoscopic techniques to treat a wide array of inflammatory and infectious disorders of the paranasal sinuses, was to expand their indications to treat other pathologies. These included cerebral spinal fluid fistulas, traumatic optic neuropathies, Grave exoophthalmopathy, refractory epistaxis, and ultimately benign and malignant tumors of the paranasal sinuses and skull base. In fact, one of the most dramatic changes in the practice of skull base surgery over the past decade has been the introduction of endoscopic techniques to approach selected tumors of the anterior cranial fossa and ventral skull base. This has been made possible by parallel advances in surgical instrumentation, hemostatic techniques and materials, image guidance systems, and most importantly collaboration between both neurosurgeons and otolaryngologists. The purpose of this chapter is to describe and highlight critical principles and philosophies that the surgeon should consider when selecting approaches to the skull base. In addition, various endoscopic approaches to the skull base from the crista galli to the odontoid will be discussed. This will be complemented by a discussion about reconstruction dilemmas and surgical complications.
MINIMAL ACCESS SKULL BASE APPROACHES The expanded endonasal approaches to the skull base can be divided into specific “modules” based on the anatomic area that needs to be accessed in addition to the extent of the tumor (9–11). This modular system is based on the control and exposure of key neurovascular structures such as the ICA and optic nerve. The modules fall under two main categories. The first includes access to the skull base from a rostral to a caudal direction, beginning at the posterior wall of the frontal sinus or crista galli and extending to the foramen magnum and odontoid. The second set of modules involves the coronal plane and represents the incremental expanded exposure and access to critical areas of the middle and posterior skull base from a medial to a lateral direction. The rostral–caudal axis involves the transcribriform, transplanum/transtuberculum, transsellar/parasellar, clivus, and odontoid/foramen magnum approaches. The coronal axis refers to exposure of structures centered on the middle third of the clivus and the key anatomic structure, which divides the coronal axis into separate modules, is the petrous ICA. The coronal plane is divided into zones based on the relationship to the petrous ICA and whether the location is suprapetrous, above the ICA or infrapetrous, below the ICA. There are two infrapetrous approaches consisting of the medial petrous apex approach (zone 1) and the petroclival approach (zone 2). The suprapetrous approaches consist of quadrangular space
PRINCIPLES AND PHILOSOPHY We believe that before one embarks on removing a tumor of the skull base, the surgeon must decide which approach best fulfills key criteria, irrespective of whether it is a conventional, external or an endoscopic approach. The first key element is that the approach should provide adequate visualization of the tumor. The second element is that the tumor be completely extirpated in adherence with oncologic principles. We believe that a complete tumor removal may be 131
132
Vescan et al.
Table 1 Summary of the Five Key Principles of Minimal Access Endoscopic Skull Base Surgery Principles of minimal access endoscopic skull base surgerya 1. Approach provides adequate visualization of tumor. 2. Tumor is removed respecting oncologic principles. 3. Preservation of major neurovascular structures. 4. Separation of sinonasal tract from intracranial contents. 5. Optimize cosmesis, minimize morbidity, and preserve quality of life. a
If one or more of these principles cannot be adhered to, an alternate approach should be considered.
approach (zone 3), superior cavernous sinus approach (zone 4), and the transpterygoid/infratemporal fossa approach (zone 5) (Fig. 1). When considering the optimal endoscopic approach for any pathology of the skull base, a detailed review of preoperative CT and MRI scans by both the otolaryngologist and the neurosurgeon is critical to plan a safe and efficient access. As a team, we routinely discuss which modules will be required, in both the coronal and the rostral–caudal plane, before each case. For the purposes of this chapter, we will discuss the most commonly employed approaches.
General Preparation: The Nasal Corridor There are certain steps and maneuvers that are common to all endonasal endoscopic approaches (EEA), irrespective of which modules will be needed for the tumor removal. Routinely, patients are positioned supine on the operating table and an orotracheal airway is secured. The head is fixed in a three-point Mayfield holder with either a slight amount of flexion (to enhance the visualization of posterior structures) or a slight amount of extension (to facilitate the visualization of anterior areas), depending on the area of the skull base that needs to be exposed. We then obtain an appropriate
image guidance registration with the aid of an optical tracking system, using surface fiducials markers that were placed prior to imaging or an LED (light emission diode)-based face mask. Once this has been completed, the nose is topically decongested with cottonoids soaked in 0.05% oxymetazoline. The middle turbinate, lateral nasal wall, and posterior nasal septum are then sequentially infiltrated with 0.5% Xylocaine, with 1:100,000 epinephrine. Resection of the middle turbinate increases the space for instrumentation and improves the visualization. Inferior turbinates are out-fractured in a two-step fashion: first by placement of blunt instrument on its undersurface displacing the turbinate medially, and second by lateral displacement of the turbinate applying pressure in a medial-to-lateral direction.
Transsellar/Transplanum This is the most commonly used skull base exposure. It is the building block for most of the extended approaches and its initial steps provide a platform and foundation for more complex and technical steps. If a vascularized septal flap is required for the reconstruction, it is harvested first to avoid devascularization and severing of the pedicle during the sphenoidotomy and posterior septectomy (see reconstruction section). Following preparation of the nasal corridor, approximately 2 cm of the posterior septum is resected to facilitate bimanual, four hand surgery. The sphenoidotomy is then widened bilaterally, using the Kerrison rongeurs. The final rectangular opening should have the lateral walls in plane with the lamina papyracea and the roof is in plane with the cribriform plate. Due to the variable pneumatization of the lateral recess of the sphenoid, it may be necessary to drill the inferolateral aspect of the sphenoidotomy over the base of the medial pterygoid plates. Superiorly, the opening is extended to the junction with the planum to allow for identification of the skull base. Inferiorly, the rostrum of the sphenoid is removed and the floor drilled posteriorly to enhance access for instrumentation. At this point, there is unencumbered visualization of the sella, medial and lateral optico-carotid recesses, clival recess, optic nerves, and ICA canals. In order to facilitate the identification of these landmarks, we remove the mucosa from the sphenoid sinus walls. This will create a moderate amount of venous bleeding, which can be controlled with warm water irrigation and the application of bone wax to high-flow areas. Paramedian septations within the sphenoid sinus require removal; however, one must be careful as these often insert directly on the carotid artery canal. A sellotomy follows removing bone from the medial aspect of the cavernous sinus on one side to medial aspect of the cavernous sinus on the other, and from the superior intercavernous sinus to inferior intercavernous sinus. This is critical to achieve an adequate sellar access for tumor removal and thus facilitates the extirpation of tumor from the lateral gutters and superoanterior areas without having to use a blind dissection technique (i.e., blind curettage). The decision of whether to remove bone over the medial opticocarotid recess or the tuberculum depends on the lateral and superior extent of tumor, respectively. Preoperative discussion between the neurosurgeon and the otolaryngologist is invaluable to plan the approach adequately.
Transcribriform Figure 1 Schematic illustration of the coronal plane containing five zones. Zone 1—medial petrous apex approach. Zone 2—petro-clival approach. Zone 3—quadrangular space approach. Zone 4—superior cavernous sinus approach. Zone 5—transpterygoid/infratemporal fossa approach.
EEA to the skull base can be divided into extradural and intradural segments. The key principle is that all dissection and visualization during the intradural portion of the case is dependent on the extradural component of the exposure. The transcribriform approach involves exposure of the anterior skull base from the posterior wall of the frontal sinus
Chapter 7: Minimally Invasive Techniques: Endonasal Endoscopic Skull Base Surgery
Figure 2 Endoscopic basal view of boundaries for an endoscopic anterior craniofacial resection. Dotted line represents the area to be resected. Source: Published with permission from Carrau RL, Kassam AB, Snyderman CH, et al. Endoscopic Transnasal Anterior Skull Base Resection for the Treatment of Sinonasal Malignancies. OperTech Otolaryngol. 2006;17:102–110.
anteriorly to the planum sphenoidale posteriorly and from the medial wall of the orbit (lamina papyracea) laterally to crista galli medially in a unilateral approach and medial orbit to medial orbit in the more commonly employed bilateral approach (Fig. 2). The nasal corridor is prepared as described above. Sometimes, it may be necessary to debulk the pedunculated aspect of the tumor first in order to be able to introduce instrumentation. The nasal septum is then disarticulated by a series of three incisions. The first is via a vertical transfixion incision at the junction of the septum and rostrum of the sphenoid. The second is an anterior vertical transfixion incision at the level of the anterior aspect of the middle turbinate or the posterior wall of the frontal sinus. This latter incision may be adapted according to the anterior extent of tumor. The two vertical incisions are then connected via a third horizontal incision parallel to the floor of the nose. By creating a septectomy defect, bimanual dissection becomes feasible. An uncinectomy, wide antrostomy, and anterior and posterior ethmoidectomy are sequentially performed. A sphenoidotomy is often performed to identify the planum sphenoidale as the posterior limit of dissection, therefore identifying the skull base. One then can proceed with the removal of all air cells in a posteroanterior direction with the skull base under direct visualization to avoid inadvertent CSF (cerebrospinal fluid) fistula. The nasofrontal recess is then widened accordingly and a Draf IIb or Draf III dissection is carried out in order to define the anterior limit of dissection (12). The lamina papyracea is then removed to provide access to the medial orbital wall. This is necessary to facilitate the control of the anterior and posterior ethmoidal arteries and serves as the lateral margin of the tumor resection. In addition, at this time the posterior wall of the maxillary sinus can be removed to access the pterygopalatine fossa. If necessary, the terminal branches of the internal maxillary artery, namely the posterior septal artery and the sphenopalatine artery, can be coagulated or clipped as needed. At this point, the mu-
133
Figure 3 Endoscopic basal view attained once osteotomies have been completed. Source: Published with permission from Carrau RL, Kassam AB, Snyderman CH, et al. Endoscopic Transnasal Anterior Skull Base Resection for the Treatment of Sinonasal Malignancies. OperTech Otolaryngol. 2006;17:102–110.
coperiosteum is stripped from both olfactory sulci leaving exposed bone. This bone is then thinned using a 3-mm hybrid cutting/diamond bur. The exposure now comprises the frontal sinus anteriorly, the planum sphenoidale posteriorly, and both orbits laterally. A rectangular area comprising the anterior skull base is then osteotomized. This comprises two paramedian osteotomies running in a rostral to caudal direction, at the junction of the orbit and the roof of the ethmoid sinuses, and two transverse osteotomies one slightly anterior to the rostrum of the sphenoid and the second rostrally, just behind the posterior wall of the frontal sinus. This allows the mobilization of the rectangular area that encompasses most of the anterior skull base (Fig. 3). It should be noted that the crista galli extends intracranially and must be thinned in an inside out fashion before being fractured or removed with a cutting punch. The intradural component of the dissection now commences with bipolar cauterization and resection of the olfactory nerves, followed by thorough cauterization of the dura. This achieves maximum vascular control outside the anterior falcine artery and controls any parasitized blood supply from the cortical circulation. Dural incisions are then performed using the previously performed osteotomies as a template and the intradural dissection proceeds from anterior to posterior. Margins are sent for frozen section to confirm histologically negative margins of resection. Histological confirmation of adequate margins is a critical and an absolute requirement during the resection of malignant tumors of the anterior cranial base regardless of the approach.
Transpterygoid/Infratemporal Fossa This approach provides access to the lateral recess of the sphenoid sinus, middle cranial fossa, pterygopalatine space, infratemporal fossa, and the lateral aspect of the middle third of the clivus for identification of the anterior genu of the petrous ICA. The nasal corridor is prepared as previously
134
Vescan et al.
rounding neurovascular structures, thus facilitating the dissection. Image guidance and specific instrumentation such as an endoscopic bipolar electrocautery and angled endoscopes are essential.
Panclival Approach
Figure 4 Axial CT Scan without contrast demonstrating prominent vidian canals leading from the pterygopalatine fossa to the apex of the petrous ICA. Drilling of the bone medial and inferior the vidian canal will allow for safe identification of the anterior genu of the ICA. Abbreviations: VC, vidian canal; ICA, internal carotid artery; Cl, Clivus.
described. The initial exposure requires an uncinectomy and a wide maxillary antrostomy that extends posteriorly until it is flush with the back wall of the antrum. Anterior and posterior ethmoidectomies and a wide sphenoidotomy widen the exposure and a posterior septectomy allows binarial and bimanual dissection techniques. In addition, it may be necessary to resect part or all of the middle turbinate for access. The crista ethmoidalis and sphenopalatine foramen can now be visualized and the posterior wall of the maxillary sinus is removed with Kerrison rongeurs. Control of the terminal vessels of the internal maxillary artery can be attained with either vascular clips or bipolar cauterization. The mucosa over the posterior wall of the sinus can be lateralized and preserved if the disease process does not involve the back wall of the maxillary sinus. Similarly, the contents of the pterygopalatine space can be lateralized to expose the base of the pterygoid plates. The vidian canal and foramen rotundum can be seen as one proceeds from medial to lateral along the back wall of the pterygopalatine fossa. Dissection can now proceed laterally to access lesions that extend through the pterygomaxillary fissure into the infratemporal fossa or posteriorly by drilling the base of the pterygoid plates. The vidian canal is a critical landmark and will lead to the second genu of the petrous ICA (Fig. 4). The vidian nerve that travels within the canal, often accompanied by the vidian artery, originates at the pterygopalatine ganglion, which is located on the superior aspect of the pterygopalatine fossa. In general, the vascular structures within the pterygopalatine space lie superficial to the neural structures. By sequentially drilling bone on the inferomedial aspect of the vidian canal, one can localize the ICA before it turns up on its vertical paraclival course. Identification of the petrous ICA is not a required component of the trans-pterygopalatine fossa approach; however, this approach is often used when further dissection and exposure necessitates identification of the artery. It should be noted that slow-growing pathologies that involve the pterygopalatine space, such as juvenile angiofibromas or schwannomas, might enlarge this corridor significantly and displace sur-
A panclival approach involves the standard initial approach that is required for a transsellar module, which will be extended inferiorly. The key principle during the clival approach is knowing the location of the petrous ICA at all times and in fact removing its bony canal in order to create the widest corridor possible for intradural dissection. Subsequent to a wide sphenoidotomy, the floor of the sphenoid sinus is drilled down until flushed with the clival recess (posterior wall of the sphenoid sinus). The inferior extent of the dissection is the level of the soft palate and the superior extent is the sella. In order to expose the lower third of the clivus, the basopharyngeal fascia needs to be dissected. This fascia is extremely resilient and a combination of techniques may be necessary for its removal. The lateral limits of the dissection are the fossae of Rosenmuller. As mentioned during the transpterygoid approach, the key element for safe identification of the anterior genu of the petrous ICA is the vidian canal and its contents. Once the vidian canal is identified in the posterosuperior aspect of the pterygopalatine, space drilling can proceed on the inferomedial aspect of the canal until it reaches the ICA at the foramen lacerum. Once this is achieved, drilling can proceed on the superior lateral aspect of the canal, thus maintaining safe depth perception of the anterior genu of the petrous ICA (Fig. 5). The exposure of the contralateral carotid artery is undertaken in the same manner followed by removal of clival bone between both paraclival carotids as they ascend vertically. During the removal of the cancellous bone within the clivus, it is common to encounter a significant amount of bleeding; this can be controlled with bone wax or, alternatively, the surgeon works through the bleeding using high-flow suction in the nondominant hand to maintain visualization. Once all of the bone work is completed, the intradural exposure is possible, thus providing access to the vertebrobasilar junction, cranial nerves V–X, pons, and the posterior circulation.
Figure 5 Intraoperative picture demonstrating the vidian nerve originating from the pterygoid wedge of the sphenoid bone. Drilling inferomedial to the vidian will allow for identification of the correct depth of the ICA.
Chapter 7: Minimally Invasive Techniques: Endonasal Endoscopic Skull Base Surgery
135
RECONSTRUCTION Skull base surgery is a rapidly evolving field with new approaches and techniques continually under development. The recent advances in endoscopic and minimal access approaches have produced complex defects that communicate the subarachnoid space and the nasal cavity. This has required ingenuity and novel techniques for the repair of these defects. There are two main options for the repair of skull base defects, the use of vascularized tissue and of nonvascularized tissue. Repair with vascularized tissue after traditional skull base approaches is well documented and widely accepted and includes regional flaps such as pericranial, galeal, and temporoparietal flaps (13–15); free flaps such as the radial forearm, latissimus dorsi, and rectus abdominis flaps (16–21). Until recently, the options for the endoscopic reconstruction of the skull base were limited to the use of nonvascularized tissue. This yielded CSF leak rates of 20–30% and became a major obstacle for the widespread acceptance of these techniques. Recently, the Haddad-Bassagasteguy flap (HBF) or nasoseptal flap has proven to be a versatile, robust vascularized flap for the reconstruction of EEA defects (22). Since its adoption, we have seen our CSF leak rates drop to numbers that compare with those of traditional techniques (around 5%). The advantage of this flap as a reconstructive option is that a second approach or incision is not necessary and the flap can be harvested endoscopically. The major drawback of the HBF is that its need must be anticipated prior to embarking on the resection as the vascular pedicle to this flap is frequently compromised during the sphenoidotomy and posterior septectomy portion of the procedure. In addition, if a revision procedure is necessary, the flap may have been used previously or the pedicle previously damaged. This leads to the second option for vascularized endoscopic skull base reconstruction, the pedicled temporo-parietal flap. The main drawback of this option is the need for a hemicoronal incision and the morbidity associated with it. However, its use in skull base surgery and other areas of head and neck surgery is well proven and its novel incorporation into endoscopic skull base reconstruction will be described below.
Vascularized Nasoseptal Flap The nose is topically decongested with 0.05% oxymetazolinesoaked cottonoids. Infiltration with 0.5% lidocaine with 1:100,000 epinephrine is undertaken, with care taken not to inject the area of the pedicle over the anterior face of the sphenoid. The flap can then be designed based on the expected size of defect. This point is counterintuitive to most reconstructive surgeons, as the flap is harvested before the resultant defect is created, thus one can be left with a flap that is inadequate. Due to this potential flaw, we tend to harvest a larger flap. The first incision is the most critical and involves use of electrocautery along the free edge of the choana from the lateral nasal wall toward the nasal septum. This incision is exceedingly difficult to perform once the flap has been elevated; extra time and patience is warranted to ensure a full incision down to the level of the sphenoid floor is performed. Two longitudinal incisions are then created from posterior to anterior. The most inferior incision is placed along the maxillary crest; however, it can be extended inferiorly and laterally to harvest the mucoperiosteum of the nasal floor. The superior incision is created 1- to 1.5-cm inferior to the cribriform plate in order to preserve olfactory function. It then crosses the rostrum of the sphenoid at level of the sphenoid sinus os.
Figure 6 Dotted lines illustrating the incisions required for harvesting a right-sided septal mucosal flap.
This leaves approximately a 1- to 2-cm width of pedicle across the anterior face of the sphenoid. Both of these incisions are joined anteriorly with a vertical incision at the level of the anterior end of the inferior turbinate (Fig. 6). Once all of the incisions are completed, the flap is elevated in a subperichondrial and subperiosteal plane back to its pedicle on the sphenopalatine foramen. The flap can then be “stored,” for the duration of the resection, in the nasopharynx or inside the antrum through a wide maxillary antrostomy. The “best” reconstruction is a constantly evolving technique. We currently employ a subdural inlay graft using a collagen matrix compound (DuraGen: Integra LifeSciences, Plainsboro, New Jersey, U.S.). This is often followed by overlay fascial acellular dermis graft (Alloderm, LifeCell Corporation, The Woodlands, Texas). Depending on the bony defect, we then lay the HBF flap directly on the denuded bone or we obliterate the dead space with a free abdominal fat graft. This is followed by fibrin glue (Tisseel, Baxter Healthcare Corporation, Deerfield, Illinois, U.S.) and Gelfoam (Johnson & Johnson, Somerville, New Jersey, U.S.). Depending on the geometry of the defect, grafts and flaps are supported with a 14 Fr Foley catheter balloon filled with saline or using an expandable sponge packing.
Transpterygoid Temporoparietal Flap The temporoparietal flap has been used extensively in head and neck reconstruction and in the repair of skull base defects. Its use as a reconstructive option during endoscopic approaches to the skull base is a novel application. The flap is based on the superficial temporal artery and its harvest necessitates a hemicoronal incision. Once the flap is harvested an endoscopic trans-antral transpterygoid approach is performed as described previously. The main reason for this approach is to create a wide tunnel for the pedicled T-P flap and to avoid any unnecessary strain on the vascular pedicle. A longitudinal incision is then fashioned in the superficial layer of the deep temporal fascia. This achieves two goals: first is to protect the frontal branch of the facial nerve and second is to gain access to the deep aspect of the zygoma for tunneling of the flap. A lateral canthotomy incision is then performed to separate the temporalis muscle from the pterygomaxillary fissure. We then employ a Seldinger technique
136
Vescan et al.
Subdural inly graft Extradural onlay graft Brain
Fat Fibrin glue
Balloon stent
struction is placed. We have found that the tissue sealant can interfere with the “take” of the grafts. In order to prevent migration of the onlay acellular dermal graft, we have used nitinol U-clips (Medtronic U-Clips, Memphis, Tennessee, U.S.) in selected situations where there is enough free dural edge to suture the graft. The application of the endoscopic sutures is a technically challenging endeavor and the nitinol U-clips are a self coiling device which we have found decreases the effort and time required for suturing. As mentioned above, the final step involves placement of either a 14 Fr Foley catheter balloon filled with approximately 10 cc of saline or expandable sponge packs. This should be done under direct visualization to avoid overinflation and compression of neural structures such as the optic nerve or shifting of the grafts.
COMPLICATIONS
Figure 7 Sagittal schematic illustrating the multilayered nonvascular reconstruction of a skull base defect. Source: Published with permission from Carrau RL, Kassam AB, Snyderman CH, et al. Endoscopic Transnasal Anterior Skull Base Resection for the Treatment of Sinonasal Malignancies. OperTech Otolaryngol. 2006;17:102–110.
using a percutaneous tracheostomy set. Concomitant endoscopic visualization is needed to confirm the passage of the guidewire into the pterygopalatine space. Once the tunnel is dilated, the external portion of the guidewire is ligated to the flap and the nasal end of the guidewire is used to pull through the flap. It is important for one surgeon or assistant to monitor the T-P flap, as it is pulled through avoiding any kinking of the pedicle. Once the flap is in the nasal cavity, it can be placed over the skull base defect and secured much the same way as with an H-B nasoseptal flap.
Nonvascularized Our technique for nonvascularized reconstruction of the skull base stems from the experience that we and others have accumulated with reconstruction of more limited defects of the skull base. The literature supports that an endoscopic repair using a variety of techniques and materials yields a success rate exceeding 95% (23–26). Our nonvascularized repair has undergone multiple changes and is constantly evolving, mainly due to the high rate of postoperative CSF fistulas we have encountered after EEA. The main premise is the use of a multilayered technique (Fig. 7). In our early experience, the first layer of reconstruction was in the epidural space between the bone and the dura. This often proved to be a technical challenge if not an impossibility due to the nature of some of the defects that were created. Our current technique involves the placement of a subdural graft of a collagen matrix compound (DuraGen: Integra LifeSciences, Plainsboro, New Jersey, U.S.). This graft is easy to manipulate due to its pliability, the goal is to overlap the dural edges by 5 mm. The next layer is an onlay graft of acellular dermis, which is placed on the bare bone. This is followed by a layer of fibrin glue. In situations where there is a need for a decrease in dead space, we use a free abdominal fat graft harvested from a periumbilical incision. An important point is that the tissue sealant should not be applied until the final layer in the multilayer recon-
Complications that result from EEA to the skull base can be divided into major and minor categories. Within the context of major complications, many are the same as those encountered during open skull base surgery; however, their incidence and severity vary as a result of the approach. Major complications include ICA and other vascular injury, CSF leak, tension pneumocephalus, cerebrovascular accidents, and intracranial infectious complications. Unique to the minor category of complications associated with EEAs are complications such as alar burns, crusting, sinusitis, dry eyes, and hypoesthesia of sensory nerves in the maxillary and mandibular distribution.
Major Complications Vascular Injuries Vascular injuries related to the expanded endonasal approach comprise capillary, venous, and arterial bleeding. Arterial bleeding can be classified as either high flow or low flow. Most of the capillary and venous bleeding that arises from bone and mucosa can be controlled with warm water irrigation. In fact, warm water irrigation (40◦ C) is used repeatedly during the case not only to stop active low-flow bleeding but also to clear the field of clotted blood which tends to absorb much of the light and affects clarity of the working picture. When focal venous bleeding is encountered from areas such as the cavernous sinus or superior and inferior intercavernous sinus, the use of microfibrillar collagen folded on a half inch by half inch cottonoid in a “sandwich” is pressed over the site to achieve hemostasis. Often, it is necessary to exchange one sandwich numerous times until venous bleeding is controlled. Arterial injury, specifically ICA injury, is the most feared complication during EEA. Injury to the ICA and its management have been described by numerous authors (27– 30). Localization and constant identification of the carotid artery is the focus of most of the dissection. Avoidance of injuries to the ICA is the best way to avoid undue morbidity to the patient. Despite extensive preoperative planning and image-guided surgery, injuries to the ICA can still occur. This is the scenario where having a two-surgeon team becomes critical. An algorithm for control of these injuries is important and all members of the surgical team should know what steps need to be instituted for prompt management. The tenets of management involve avoiding hypoperfusion of the brain, while at the same time correcting hypovolemia and maintaining controlled hypertension for brain perfusion. Oxygen (100%) can also be administered to optimize oxygenation. During this initial phase, communication between the skull
Chapter 7: Minimally Invasive Techniques: Endonasal Endoscopic Skull Base Surgery
base surgeon and the anesthesiologist is paramount. For most injuries to the ICA during endoscopic skull base surgery, the safest method of controlling the vessel is probably sacrifice of the vessel via angiography and embolization with internal coiling. Before one can proceed to the interventional radiology suite, the hemorrhage must be controlled. This can be done through a variety of techniques: packing, muscle and fascia onlay grafting, and clipping and bipolar cautery(31). It is important to do a follow-up angiogram 5 to 14 days following the injury to rule out formation of a pseudo aneurysm.
Cerebrospinal Fluid Leak CSF leak is the most commonly encountered complication of EEA approaches to the skull base. With the recent introduction of the H-B nasoseptal flap, we have seen dramatic reductions in the incidence of CSF fistulas. There are numerous factors that predispose to development of a CSF leak following endoscopic skull base surgery. These include, size of the defect, extent of intracranial dissection, opening of an arachnoid cistern or ventricle, prior radiation therapy, patient body mass index, and, ultimately and most importantly, the technique of repair. The diagnosis of a CSF leak is a combination of clinical findings of clear rhinorrhea and a positive “halo” sign in addition to a positive beta-2 transferrin test. At our institution, we are able to obtain a beta-2 test in less than 24 hours; however, this is not always possible. Routine use of lumbar drainage is not necessary and we recommend selectively employing lumbar drainage in patients with elevated CSF pressures or in those patients that an arachnoid cistern or ventricle was opened (25). We found that once the leak has been documented, the majority of these patients will require an exploration and repair of the leak. The patient is taken to the operating room, no preoperative localization studies are performed, and the reconstruction is taken down until the site of leak is identified. Once this is done, we then employ a variety of reconstructive techniques depending on vascularized tissue available and size of the defect, as described previously in this chapter. In a small percentage of patients, conservative treatment with a lumbar drain and the usual CSF leak precautions achieves a complete closure. Tension Pneumocephalus This is an uncommon complication following endoscopic approaches to the skull base. There are numerous theories as to why this is not as common during endoscopic approaches as compared with open approaches. One of them being that we tend to completely occlude the posterior nasopharynx, thus preventing any airflow to the skull base with our reconstructive techniques, either using a Foley catheter or a merocel pack. Another theory is that the reconstruction involves some sort of overlay technique from below which can act as a physical barrier and prevent any ball valving of the skull base. The management of tension pneumocephalus after an EEA involves the removal of all packing materials and potentially aspiration of air via a burr hole, needle, or ventriculostomy, in addition to the administration of 100% oxygen. Close neurologic monitoring is necessary.
Infectious Complications The main infectious complication to be concerned with is bacterial meningitis. Our incidence of culture proven bacterial meningitis is very low and is a mirror image of our open skull base experience. Ascending meningitis can occur secondary to direct spread via a breakdown in the reconstruction; it can also occur secondary to extensive crusting and sinusitis in the absence of a CSF fistula. Active bacterial sinusitis associated
137
with mucopurulent secretions is a contraindication to proceeding with intradural work. If this is encountered during the approach, the operation must be staged with the patient receiving treatment for sinusitis in the interim. The meningitis can be treated with culture-directed antibiotics in addition to reversal of the underlying etiology.
FUTURE DIRECTIONS Since the inception of skull base surgery as a discipline in the 1960s, dramatic changes have occurred with respect to the surgical management of the wide variety of pathologies that can be found in this anatomic area. The advancement of the field has been directly related to collaboration both between skull base surgeons and between physicians and the biomedical industry that supports them. Future directions include advancement of imaging technology such that it will be possible to acquire fast and precise imaging during the actual procedure. Current image guidance technology is based on a preoperative scan and does not account for tissue shift or surgical resection during the procedure. The advent of fast and portable CT scanning will allay many of these problems. In addition, the endoscopic skull base surgeon is inundated with information and data during the procedure. Future developments would consist of new ergonomically designed displays and information feedback mechanisms that would allow seamless synchronization of data for the primary surgeon. It is also the responsibility of the leading endoscopic and other minimal access skull base surgeons to provide good outcomes data to justify a variation from the traditional methods. Despite clear improvement in patient morbidity with minimal access procedures, what remain to be defined are the long-term outcomes. This will be a key element to forwarding the specialty. Future development and implementation of robotassisted surgery could aid in technically challenging maneuvers. This will complement new biological glues and chemotherapeutic delivery systems. As with the introduction of any new procedures, training of future generations will be a requisite. Surgical simulators and novel teaching techniques will be instrumental in acquiring minimal access skill sets to shorten the learning curve for these approaches. In summary, the future of minimal access skull base surgery is exciting. There are many areas for advancement and refinement and the future is full of opportunity.
REFERENCES 1. Ketcham AS, Wilkins RH, Van Buren JM, et al. A combined intracranial facial approach to the paranasal sinuses. Am J Surg. 1963;106:698–703. 2. Cohen-Gadol AA, Liu JK, Laws ER. Cushing’s first case of transphenoidal surgery: The launch of the pituitary surgery era. J Neurosurg. 2005;103(3):570–574. 3. Carrau RL, Kassam AB, Snyderman CH. Pituitary Surgery. Otolaryngol Clin North Am. 2001;34(6):1143–1155. 4. Hardy J. Transsphenoidal hypophysectomy. J Neurosurg. 1971;34(4):582–594. 5. Stammberger H, Posawetz W. Functional endoscopic sinus surgery. Concept, indications and results of the Messerklinger technique. Eur Arch OtorhinoLaryngol. 1990;247(2):63–76. 6. Messerklinger W. Background and evolution of endoscopic sinus surgery. Ear Nose Throat J. 1994;73(7):449–450.
138
Vescan et al.
7. Hosemann W, Gode U, Wigand ME. Indications, technique and results of endonasal endoscopic ethmoidectomy. Acta Otorhinolaryngol Belg. 1993;47(1):73–83. 8. Kennedy DW. Technical innovations and the evolution of endoscopic sinus surgery. Ann Otol Rhinol Laryngol Soppl. 2006;196:3–12. 9. Kassam A, Carrau RL, Snyderman CH, et al. Expanded endonasal approach: Fully endoscopic, completely transnasal approach to the middle third of the clivus, petrous bone, middle cranial fossa and infratemporal fossa. Neurosurg Focus. 2005;19(1): E6. 10. Kassam A, Snyderman CH, Carrau RL, et al. Expanded endonasal approach: The rostrocaudal axis. Part I. Crista galli to the sella turcica. Neurosurg Focus. 2005;19(1):E3. 11. Kassam A, Snyderman CH, Carrau RL, et al. Expanded endonasal approach: The rostrocaudal axis. Part II. Posterior clinoids to the foramen magnum. Neurosurg Focus. 2005;19(1):E4. 12. Weber R, Draf W, Kratzsch B, et al. Modern concepts of frontal sinus surgery. Laryngoscope. 2001;111(1):137–146. 13. Snyderman CH, Janecka IP, Sekhar LN, et al. Anterior cranial base reconstruction: Role of galeal and pericranial flaps. Laryngoscope. 1990;100:607–614. 14. Neligan PC, Mulholland S, Irish J, et al. Flap selection in cranial base reconstruction. Plast Reconstr Surg. 1996;98:1159–1166. 15. Urken ML, Catalano PJ, Sen C, et al. Free tissue transfer for skull base reconstruction: Analysis of complications and a classification scheme for defining skull base defects. Arch Otolaryngol Head Neck Surg. 1993;119:1318–1325. 16. Funk GF, Laurenzo JF, Valentino, et al. Free tissue transfer reconstruction of midfacial and cranio-orbital-facial defects. Arch Otolaryngol Head Neck Surg. 1995;121:293–303. 17. Izquierdo R, Leonetti JP, Origitano TC, et al. Refinements using free-tissue transfer for complex cranial base reconstruction. Plast Reconstr Surg. 1993;92:567–575. 18. Jackson IT, Adham MN, Marsh WR, et al. Use of galeal frontalis myofascial flap in craniofacial surgery. Plast Reconstr Surg. 1986;77:905–910.
19. Jackson IT, Marsh WR, Hide TA. Treatment of tumors involving the anterior cranial fossa. Head Neck Surg. 1984;6:901–913. 20. Johns ME, Winn HR, McLean WC, et al. Pericranial flap for the closure of defects of craniofacial resection. Laryngoscope. 1981;91:952–959. 21. Jones NF, Schramm VL, Sekhar LN. Reconstruction of the cranial base following tumor resection. Br J Plast Surg. 1987;40:155–162. 22. Hadad G, Bassagasteguy L, Carrau RL, et al. A novel reconstructive technique after endoscopic expanded endonasal approaches: Vascular pedicle nasoseptal flap. Laryngoscope. 2006;116:1882– 1886. 23. Zweig JL, Carrau RL, Celin SE, et al. Endoscopic repair of CSF leak to the sinonasal tract: Predictors of success. Otolaryngol Head Neck Surg. 2000;123:195–201. 24. Hegazay HM, Carrau RL, Snyderman CH, et al. Transnasal endoscopic repair of cerebrospinal fluid rhinorrhea: A meta-analysis. Laryngoscope. 2000;110:1166–1172. 25. Carrau RL, Snyderman CH, Kassam AB. The management of the CSF leaks in patients at risk for high pressure hydrocephalus. Laryngoscope. 2005;115:205–212. 26. Zweig JL, Carrau RL, Celin SE, et al. Endoscopic repair of acquired encephaloceles, meningoceles, and meningoencephaloceles: Predictors of success. Skull Base Surg. 2002;12:133–139. 27. Oeken J, Bootz F. Severe complications after endonasal sinus surgery: An unresolved problem. HNO. 2004;52:549–553. 28. Weber R, Draf W. Complications of endonasal microendoscopic ethmoid bone operation. HNO. 1992;40:170–175. 29. Rauchfuss A. Complications of endonasal surgery of the paranasal sinuses. Special anatomy, pathomechanisms and surgical management. HNO. 1990;38:309–316. 30. Kassam A, Snyderman CH, Carrau RL. Endoneurosurgical hemostasis techniques: Lessons learned from 400 cases. Neurosurg Focus. 2005;19(1):E7. 31. Weidenbecher M, Huk WJ, Iro H. Internal carotid artery injury during functional endoscopic sinus surgery and its management. Eur Arch Otorhinolaryngol. 2005;262:640–645.
8 Reconstruction of Skull Base Defects Peter C. Neligan, Christine B. Novak, and Patrick J. Gullane
The pedicled myocutaneous flap, including the pectoralis major and the latissimus dorsi flaps, introduced another reconstructive option for patients following tumor ablation in the head and neck region. Ariyan (20), in 1979, used the pectoralis major flap and then described a variation that included externalizing the flap to repair defects as far as the orbital region (21). Quillen et al. (22), in 1978, described the latissimus dorsi myocutaneous flap for reconstruction of head and neck defects. This flap provided a number of advantages over previously described flaps, including a large arc of rotation, which allowed the flap to be extended to the floor of the middle fossa. However, because of the anatomic location of the latissimus dorsi muscle, the patient must be repositioned to elevate this flap. In general, this myocutaneous flap is rarely used for primary reconstruction in the skull base. The free tissue transfer has transformed and improved skull base reconstruction. The free flap provides an abundant source of well-vascularized tissue for repair of large cranial base defects (23). Unlike pedicled flaps, design and placement are not limited by the attachment of the pedicle and the arc of rotation and therefore dependent on the requirements of the reconstruction multiple flap designs may be considered. Free tissue transfer also provides the option of using two surgical teams which can decrease anesthesia time by having one team continue with tumor ablation while the second team proceeds with the flap harvest. While excellent patient outcome depends on many factors including microvascular expertise and correct patient selection, success with free tissue transfer is reported to be greater than 95%. The multidisciplinary surgical team has expanded the surgical options for treatment of skull base tumors by providing expertise from the head and neck surgeon, the plastic reconstructive surgeon and the neurosurgeon.
INTRODUCTION Lesions located in the skull base create a challenge for the reconstructive surgeon. Ablation of neoplasms in this region often requires extensive resection and this results in a large defect that often requires complex reconstruction. Because of the anatomic location of the skull base and the potential for postoperative complications, this region has a unique complexity that requires a comprehensive approach. The management of patients with skull base tumors has advanced with the improvement of operative technique, diagnostic and interventional radiology, and the introduction of the multidisciplinary surgical and medical teams. This has provided the opportunity for successful surgical treatment of most cranial base tumors, which were previously deemed inoperable.
HISTORY Crile (1), in 1906, introduced the radical neck dissection and prior to the use of this approach the prognosis was poor for patients with head and neck cancer who were treated with surgery (2). Early in the 20th century, only a limited number of reconstructive options were available and therefore skin grafts were commonly used for coverage and these grafts were often put directly on to the dura or cranial bone (3– 5). This resulted in many reconstruction failures. In patients where reconstruction of the orbit was required, tubed pedicles were used in multiple staged procedures (6). Golovine (7)and then Gillies (8) popularized the temporalis muscle flap for soft tissue orbital reconstruction; the temporalis flap remains a popular reconstructive option for other defects in the skull base region (6,9–12). For orbital reconstruction, Thomson (13) described using the ipsilateral forehead flap. Expanding the use of the forehead flap, McGregor described this flap for reconstruction of intraoral lining (14). In 1965, Bakamjian (15) described the deltopectoral flap for oropharyngeal reconstruction. This was followed by a modification by McGregor and Jackson (16) that allowed the flap to be lengthened so that it could be extended to the ear. Fisch et al. (17) broadened the use of this flap by suggesting that the flap could be internally tunneled to reconstruct the nasopharynx and also if the flap was transferred externally, it could be extended to the orbital and zygomatic regions. Historically, the temporalis and deltopectoral flaps were the workhorse flaps for reconstruction in the head and neck region (2). While these flaps do provide good soft tissue coverage, the donor site morbidity is considerable. With defects that are located superior to the palate, the reported complication rate is between 40% and 50% when the deltopectoral flap is used (18,19) and this is likely due to the failure of these flaps to provide a watertight seal of the oral cavity.
ANATOMY The intracranial surface of the skull base is created by the floor of the posterior, middle, and anterior cranial fossae, and the extracranial component provides structure to the orbital roof, sphenoid sinus, nasopharynx, and infratemporal fossa. This area of the cranial base has many critical anatomic structures that transverse the cranium via foramina and canals. The unique function and complex location of the skull base makes the management of any pathologic lesion in this region challenging. To assist in the evaluation and treatment of skull base lesions and because of the complexity of the cranial base, this area has been classified into different regions or zones (17,24,25). The skull base was divided into the anterior and posterior areas by Jackson and Hide (17) to describe the 139
140
Neligan et al.
Skull base defect
Region I
Small size defect
Watertight dural seal
Region II
Local vascularized flap
Large size defect
Watertight dural seal
Free tissue transfer
Region III Midline zone 1 defect
Figure 1 The skull base is divided into 3 regions based upon the anatomic location and growth pattern of the tumors. Source: Reproduced with permission from Ref. 23.
requirements for operative reconstruction; the anterior area is associated with the anterior cranial fossa and the posterior area was subdivided into the posteroanterior, posterior– central, and posterior–posterior segments. Jones et al. (25) separated the skull base into the anterior, middle, and posterior regions, which corresponded to the anterior, middle, and posterior cranial fossae. Irish et al. (24) reviewed 77 patients with skull base neoplasms and divided the skull base into three regions (Fig. 1) based upon the anatomic boundaries and tumor growth patterns within the different regions of the skull base. As described by Irish et al. (24), Region I tumors arise from the sinuses, orbit, and other anterior structures and may involve lesions extending into the anterior cranial fossae. In addition, Region I tumors include those lesions that originate from the clivus and those that extend posteriorly to the foramen magnum. These tumors were included in Region I because of the similarity to other Region I tumors and also they are surgically accessed using an anterior approach. Tumors that are classified in Region II originate in the lateral skull base and may have extension to the middle cranial region but primarily involve the infratemporal and pterygopalatine fossa. Region III lesions occur in the area around the parotid and temporal bone and ear and these tumors extend intracranially to involve the posterior cranial fossae (24). The patient sample evaluated by Irish et al. (24) included only patients with neoplasms that entered the skull base and did not include tumors that projected only to the skull base. Outcome data reported in other studies vary significantly because many reports include patients with extracranial tumors and tumors with bony invasion. One of the difficulties when comparing outcome in patients with skull base neoplasms is the inclusion of patients with tumors in different locations within the skull base and varying pathologies. Skull base tumors for this review are defined as those neoplasms that require both an intracranial and an extracranial approach for ablation.
RECONSTRUCTION OF THE SKULL BASE In general, head and neck reconstruction requires consideration of postoperative function and cosmesis and must pro-
Galea-frontalis flap Pericranial flap
Lateral zone 1 defect
Zone 2 or zone 3 defect
Galea –frontalis flap Temporalis flap Pericranial flap Regional pedicled flap Temporalis flap e.g., trapezius Temporoparietal fascial flap
Figure 2 Algorithm for management of skull base defects. Source: Reproduced with permission from Ref. 61.
vide structural support and soft tissue bulk. The additional unique challenges of reconstruction of skull base defects is the need to ensure a watertight dural seal, to fill dead space, and to provide coverage with well-vascularized tissue. Adherence to these reconstructive goals will provide the best opportunity for optimal patient outcome with the least risk of postoperative complications. The specific reconstructive procedure that is required will be selected based upon several key factors, including the exact location of the defect, the defect size, the tissue involved, and if there has been exposure of the dura (Fig. 2). In this review of reconstruction of skull base defects, we have utilized the regional classification system based on the classification described by Irish et al. (24).
Skull Base Zone I Zone I is compatible with the anterior skull base defects described by Jackson and Hide (17) and Jones et al. (25) and extends from the midline anteriorly to the posterior wall of the orbit and also includes an extension to the foramen magnum. Zone I tumors originate in the anterior sinuses, midfacial skin, and also may extend to the cribriform plate, the dura and brain tissue. In the patients described by Irish et al. (24), the most common location of the tumors was in Zone I (44%) and 6% of patients had lesions in Zone I and II. Surgery in the skull base requires a direct operative approach, which will provide access to the entire tumor. While the location of the tumor is important in selection of the surgical approach, consideration must also be given to the size and characteristics of the lesion. In Zone I, a lateral rhinotomy or maxillotomy approach is most commonly used and in some cases a partial or total maxillectomy may be necessary. Tumors extending into the region of the clivus can be accessed by combining a Le Fort I maxillary osteotomy and mandibulotomy with midline split of the soft and hard palates (26). To provide complete exposure for tumor excision and to protect vital neural and vascular structures, a bicoronal incision and frontal craniotomy may also be necessary.
Chapter 8: Reconstruction of Skull Base Defects
(A)
(B)
(E)
(C)
141
(D)
(F)
Figure 3 (A–F) (A) Patient is seen preoperatively with proptosis due to a Region I tumor. (B) The anterior skull base tumor is seen on MRI. (C) The rectus abdominus free tissue transfer is prepared. (D) The flap is inset. (E, F) The patient is seen with excellent outcome six years following surgery. Source: Reproduced with permission from Patrick J. Gullane, MB & Peter C. Neligan, MB.
Following tumor ablation, the reconstructive options available will be determined by the type of tissue required, the size and the position of the defect, and if there was exposure of the dura. Preceding the development of myocutaneous flaps, the reconstruction consisted of coverage using split thickness skin grafts and this was utilized even in patients where the dura was exposed. The use of skin grafts in many cases was unsuccessful and similarly, when tensor fascia lata grafts were used to repair the dura, these efforts proved ineffective. Ketcham et al. (3) in their review, found that approximately 50% of these patients developed cerebrospinal fluid leaks. Because of the failed experience of using a nonvascularized skin graft to cover a nonvascularized tensor fascia lata graft, vascularized tissue was introduced for these repairs. There are numerous local soft tissue flaps which have been reported for use with skull base defects, including forehead flaps (13,27,28) and glabellar flaps (29). In patients where a bicoronal incision was necessary, the pericranial flap (30,31) and the galeal flap (32,33) have also been described. The pericranial and galeal flaps may be used when there is a small midline defect and they provide good soft tissue coverage with minimal donor site morbidity. In patients with defects that are located more laterally in the region of the anterior cranial fossa, the temporalis muscle can be used. This flap is not as dependable when it is medially transposed, due to decreased vascularity. The most distal portion of the flap is most vulnerable to vascular compromise and it is often this distal area that is vital for successful closure. Therefore, if there is necrosis of the flap and/or wound dehiscence in this region, the risk of failure of the reconstruction increases. Donor site morbidity and patient dissatisfaction may also be a problem with the temporalis flap. With transposition of the temporalis muscle, a concave deformity will occur at the donor site and this may not be acceptable to all patients. If the temporalis muscle has been devascularized by tumor ablation or neck dissection, then this muscle may not be available
and other sources of soft tissue must be considered for the reconstruction. Myocutaneous or muscle flaps provide vascularized soft tissue bulk and this is helpful to cover an area that may require postoperative radiation or may have been irradiated preoperatively. These muscle flaps provide good soft tissue contour and can minimize the risk of postoperative complications by filling dead space in the skull base region. The trapezius, the latissimus dorsi, and the pectoralis major muscles have been used as pedicled myocutaneous flaps for the management of these defects (34). Rosen (35) described the extended trapezius flap to reconstruct defects that are located more anterior in the skull base. Although pedicled flaps can reach the defect in many cases and have been historically used for reconstruction in this region, they have been become obsolete in favor of free tissue transfers and are no longer the first choice for skull base reconstruction. However, pedicled flaps may be used as a secondary procedure in patients who have had a failure of a free flap reconstruction. When large defects remain following tumor ablation, free tissue transfer is the best reconstructive option for the majority of patients (34,36–38). It provides well-vascularized tissue and because a free flap does not have the attachment of a pedicle, it can be designed and placed in the desired position and it provides well-vascularized tissue. With a pedicled flap, the design is such that the most distal portion of the flap is usually the aspect that is essential for successful soft tissue coverage. In most pedicled flaps, this distal region is the least reliable in terms of vascularity and therefore in cases where distal flap necrosis occurs, the whole reconstruction may fail. Free tissue transfer flaps can be designed to provide well-vascularized tissue in the best position to effectively seal the defect (Fig. 3A–3F). Because a microvascular free tissue transfer may be harvested at the same time as the tumor ablation, it is possible to utilize a two-team approach; one team performing the tumor ablation and the second team
142
Neligan et al.
performing the flap harvest. This two-team approach minimizes patient anesthesia time and thus decreases the risk of anesthesia-related complications. When the dura and aerodigestive tracts have been exposed during tumor ablation, it is necessary to provide a barrier between these two tracts, and myocutaneous flaps are particularly useful in this role because of their capability for resistance of infection. Numerous muscle flaps have been described for soft tissue reconstruction and the rectus abdominis myocutaneous flap is often used to repair large defects in this area of the cranial base (34,39). It is a well-vascularized flap that can provide good soft tissue coverage for large defects and thereby provides an excellent barrier between the aerodigestive and dura tracts. In Zone I, reconstruction of small defects may be accomplished with a local vascularized flap, such as the pericranial or galeo-frontalis flap and this type of flap will provide sufficient soft tissue. When a large muscle flap is required in cases of more extensive Zone I defects, the rectus abdominous flap is the flap of choice in our unit.
topectoral flap (42) have been used for repair of soft tissue defects located in Zone II. For smaller defects that are located in a more lateral region, a temporalis muscle flap may be used. Other muscle flaps including the pectoralis major flap and the trapezius flap have also been described for reconstruction in Zone II defects. These flaps have not proven to be consistently reliable and therefore are used less often. Other alternatives include a free tissue transfer such as the rectus abdominis muscle, which provides ample soft tissue bulk, a long pedicle and reliable vascularity. The rectus abdominus muscle flap can also be designed to provide muscle extensions. When necessary, these extensions are useful to obliterate the sphenoid sinus and to provide coverage in the neck region, particularly when a saphenous vein graft has been used to reconstruct the carotid artery. The use of the free tissue transfer also provides the opportunity to utilize separate surgical teams for the tumor extirpation and for the reconstruction, which will significantly reduce operative time.
Skull Base Zone III Skull Base Zone II Zone II (24) is formed by the infratemporal and pterygomaxillary fossae and a segment of the middle cranial fossa and it extends from the posterior aspect of the petrous temporal bone to the posterior orbital wall. There are several neural and vascular structures that traverse this zone to the middle cranial fossa. The foramen lacerum is positioned on the skull’s under surface and the internal carotid artery passes through this foramen to the carotid groove on the sphenoid bone. The mandibular and maxillary branches of the trigeminal nerve are also found in this zone; the mandibular nerve traverses through the foramen ovale and the maxillary nerve through the foramen rotundum. The facial nerve (cranial nerve VII) and the auditory nerve (cranial nerve VIII) pass through the petrous temporal bone. Neoplasms, which originate within Zone II, include nasopharyngeal tumors, glomus jugulare neoplasms, clival chordomas, and meningiomas. Common tumors that occur external to Zone II but extend into the middle cranial base include basal and squamous cell carcinomas of the external and middle ear and scalp and parotid tumors. In the review by Irish et al., only 9% of the patients had tumors that were isolated to Zone II and 43% of these neoplasms had dural involvement (24). As compared to Zone I and III, tumors involving Zone II are not as common but are associated with the worst prognosis (24). Tumors that are located within the middle skull base region may be surgically removed through an infratemporal approach (40) using a hemicoronal incision and preauricular extension. This approach provides surgical access to the mandibular condyle and if necessary allows inferior displacement of the zygomatic arch following dissection of the facial nerve from the stylomastoid foramen to the parotid gland. If an increased surgical exposure is necessary, the infratemporal approach may be combined with a mandibulectomy, a lateral mandibulotomy or an anterior mandibulotomy with anterior displacement of the mandible. A transtemporal approach (41) has also been utilized for ablation of neoplasms within Zone II. This approach necessitates a hemicoronal incision with a postauricular extension with transection of the external auditory canal. If exposure of the intradural component of middle cranial fossa is necessary, a frontotemporal craniotomy can also be used. The reconstructive options that are available will vary and are based upon the size and location of the defect following tumor ablation (Fig. 2) and patient factors. Historically, pedicled flaps such as a rotational scalp flap (28) and del-
Zone III (24) consists primarily of the posterior cranial fossa and includes some of the posterior segment of the middle cranial fossa. This zone includes the internal jugular vein, the glossopharyngeal nerve, the vagus nerve, and the accessory nerve as they traverse the jugular foramen from the posterior cranial fossa. The hypoglossal nerve is also in Zone III but passes through the anterior condylar canal. As reported by Jones et al. (25), the most common tumors occurring in Zone III were schwannomas and glomus tumors. The occurrence of this type of neoplasm was not consistent with the study by Irish et al. (24), where they noted 14 of 25 Zone III tumors were squamous cell carcinomas. The variation in the pathology reported in these two studies may signify a difference in the patients that are referred to different surgical specialties. Zone III tumors are commonly exposed using a transtemporal approach and in some cases bony resection around the carotid artery and sigmoid sinus is necessary. Following tumor ablation, small defects may be reconstructed using local muscle flaps such as the temporalis or the sternocleidomastoid muscle or the temporoparietal pedicled flap. If a radical neck dissection is necessary, then in many cases these muscles will be devascularized or excised in the resection and therefore the local muscles are not available for reconstruction. The latissimus dorsi muscle flap has been described for use in Zone III defects. However, the use of this flap requires that the patient be repositioned for flap harvest. Also the latissimus dorsi muscle flap is not as reliable as other free tissue transfers and therefore this flap is not the flap of choice for use with primary reconstruction. In patients where a large defect remains following tumor excision, a free tissue muscle flap such as the rectus abdominis is our choice. This flap provides a well-vascularized soft tissue coverage and is very reliable for reconstruction of complex defects in this region (Fig. 4A–4G ). The variety of pathologies of skull base lesions and variation in the exact location contribute to the difficulties in interpreting studies that have reported outcome and patient prognosis. It appears that squamous cell carcinomas are the most frequently reported tumors in the skull base (24,43). Irish et al. (24) reported that 29% of their patients had a diagnosis of squamous cell cancer and these tumors were most commonly located in Regions I and III. However, basal cell carcinomas, chordomas, chondrosarcomas, and esthesioneuroblastomas were most frequently found in Region I (24). Tumors that invaded both Regions II and III usually originated in the salivary gland and typically went from
Chapter 8: Reconstruction of Skull Base Defects
(A)
(B)
(E)
(C)
(F)
143
(D)
(G)
Figure 4 (A–G)(A) Patient is seen preoperatively with large mass. (B) Region III squamous cell carcinoma is shown on MRI. Source: Reproduced with permission from Patrick J. Gullane, MB & Peter C. Neligan, MB. (C) Region III squamous cell carcinoma is shown on CT scan. Source: Reproduced with permission from Ref. 61. (D) The area of resection, including the dural patch, is illustrated. Source: Reproduced with permission from Ref. 61. (E) Preparation of the area to be resected. Source: Reproduced with permission from Ref. 61. (F) Resected specimen. Source: Reproduced with permission from Ref. 61. (G) Patient presents postoperatively following a reconstruction using a rectus abdominis myocutaneous free tissue transfer. Source: Reproduced with permission from Ref. 61.
Region II into Region III (24). In the series reported by Irish et al. (24), there was better survival at two and four years in patients with Region I and III tumors and no patients with Region II tumors survived for four years (24). Prognosis not only depends on tumor pathology but also depends on patient comorbidities. In the collaborative review by Patel et al. (43), the extent of intracranial involvement and comorbid medical conditions significantly influenced patient survival.
TECHNICAL CONSIDERATIONS Because of the location of the skull base and vital anatomic structures that are located in this region, it is important to minimize the risk of complications because of the potential for life-threatening consequences. Therefore, there are aspects of skull base reconstruction that necessitate special attention and meticulous technique to avoid any preventable complications. It is important to appreciate the goals of the reconstruction and then the technical considerations become more evident. Obliteration of the dead space is essential for a successful reconstruction and therefore muscle and myocutaneous flaps are often utilized. Muscle is well vascularized and very malleable and therefore it can be packed into cavities and fills spaces within the skull base. In cases where the dura has been breached, successful reconstruction also depends on a watertight dural seal. The anatomy of the skull base is such that there is a downward gravitational pull on any reconstructive
flaps. This likely contributes to the separation of the dural repair and therefore makes it difficult to maintain a tight seal of the repair and thus creates more dead space. To solve this problem and to maintain the dural seal, the flap can be maintained in position by attaching it to the surrounding bony structures. It is secured by drilling holes in the bone and the sutures are placed through these holes (Fig. 5A and 5B). The flap may then be pulled up into the defect and further secured with fibrin glue. This will provide a more effective watertight seal between the flap and the dura. Reconstructions that utilize a free tissue transfer require optimal donor blood vessels to increase the likelihood of a successful reconstruction. The superficial temporal vessels are reliable donor vessels for most Zone I reconstructions (44,45). In cases where the superficial temporal vessels do not appear to be optimal or are not available, the pedicle can be extended to the neck and the facial or superior thyroid vessels may be used. Reconstruction in Zones II and III often utilizes vessels located in the neck and when selecting these vessels, consideration is given to several factors including the vessel caliber, pedicle length, and pedicle geometry. In most cases, vein grafts are unnecessary. Because the location of the skull base is not superficial, the reconstructive flap is often deep and therefore clinical monitoring of perfusion is not possible. In these patients, venous pedicle patency may be monitored with an implantable Doppler probe. In patients who are treated with postoperative radiation therapy, there is an increased risk of osteoradionecrosis or
144
Neligan et al.
Similar to bone, appropriate vascularized soft tissue coverage must also be included in the reconstruction to minimize the risk of postoperative complications (46).
COMPLICATIONS Because of the location of the skull base and the close proximity of vital neural and vascular structures, postoperative complications can quickly develop into life-threatening conditions. Therefore, it is essential to minimize the risk of postoperative complications and this is achieved by selecting the most optimal reconstructive procedure based upon the site and size of the surgical defect and associated patient factors. The complication rate following skull base reconstruction that has been reported in previous studies varies from 11.5% to 63% and there has been a decrease in the rate in the more recent patient studies (43,47). There are several factors that have likely contributed to the decrease in complications and these include improved surgical technique, enhanced postoperative patient care, and the use of the free tissue transfer. Complications that occur in the early perioperative period may compromise the reconstruction. These early complications include infection, dural exposure, cerebrospinal fluid leak, and ascending meningitis. The occurrence of the majority of the early complications is usually related to a poor dural repair and/or persistent communication between the dura and the aerodigestive tracts. Late complications are not life threatening but will certainly impact on patient morbidity, satisfaction, and quality of life. Late complications are often the result of a lack of structural support, soft tissue atrophy, or postoperative radiation fibrosis.
(A)
Early Complications Wound Infection The most common early complication is wound infection (43). Because of the location of the skull base, this is a potentially fatal complication that requires prompt treatment. In cases where the oral and/or nasal cavities have been breached, prophylactic antibiotics such as cephalosporin with an anaerobic agent may be instituted. Otherwise, treatment of a wound infection should include wound cultures and based on the laboratory results, the appropriated antibiotic treatment can be commenced.
Flap Compromise (B)
Figure 5 (A) To facilitate suspension of the flap, suspension sutures are passed through drill holes in the bone following resection of the temporal bone. (B) The rectus abdominus muscle free tissue transfer (within white line) is inset and it is suspended from the temporal bone defect using suspension sutures. Source: Reproduced with permission from Ref. 61.
infection when avascular bone is used in the reconstruction. Therefore, the use of avascular bone should be avoided and in most cases, a soft tissue reconstruction is preferential to bony reconstruction in the skull base region. When bony reconstruction is deemed to be necessary, avascular bone should not be used and vascularized bone should be selected. The scapular graft is a good choice to be used for the donor. A simpler alternative, such as titanium mesh, may also be used to provide structural support (46). If allograft material is selected, then titanium is an excellent choice because it is an inert nonallergenic metal and follow-up radiologic studies, including CT and/or MR imaging are not contraindicated.
As reported by Califano et al. (36), there are more free tissue transfers now used for reconstruction of skull base defects. Compared to a historical cohort who did not have free flaps, the authors reported a similar complication rate in patients who had free flaps, although these patients involved more complex resections and reconstructions. Neligan et al. (34) reported a decrease in postoperative complications in patients with free tissue transfers compared to patients with pedicled flaps. In those patients who had a free flap, there were only 10% of cases with compromised wound healing and 5% of patients had a cerebrospinal fluid leak (34). With improved surgical technique and perioperative care, the complications associated with free tissue transfers have decreased and therefore this is the reconstructive option recommended for most patients with large skull base defects following tumor ablation.
Cranial Nerve Dysfunction Because of the location of the skull base and the close proximity of the cranial nerves, postoperative dysfunction in one or more cranial nerves is not unexpected. In cases where the
Chapter 8: Reconstruction of Skull Base Defects
cranial nerve is included in the resection of the neoplasm, reconstruction of the nerve using direct nerve coaptation or a nerve graft should be attempted. In patients where the nerve has not been resected or transected, patients should be followed for reinnervation and appropriate intervention should be instituted. In patients with skull base pathology, dysfunction of the muscles innervated by the facial nerve may occur from tumor pathology, growth, ablation, or operative trauma. Injury to the facial nerve can result in significant morbidity associated with decreased self-image, function, and socialization. Therefore, facial nerve reconstruction to achieve symmetry and/or reanimation is recommended to minimize patient morbidity and to improve patient outcome. Although a complete facial palsy may be present, the absence of facial movement does not indicate the severity or degree of facial nerve injury. A neurapraxic (48) or a Sunderland first degree (49) injury may occur as a result of operative trauma, such as retraction and complete recovery would be anticipated following neural remyelination. Increased neural trauma may cause a more severe degree of nerve injury; however, reinnervation of the facial muscles and complete recovery may still occur following an adequate time for nerve regeneration (1 mm/day) (49). Because the nerve is intact and the nerve injury is not severe enough to preclude nerve regeneration, these patients will not require any operative intervention to regain facial function. In patients with no evidence of recovery, postoperative electromyography (EMG) should be performed approximately four weeks following surgery to determine the severity of nerve injury and the potential for recovery. EMGs performed before this time may not adequately assess the degree of nerve injury. Because of the inability for complete eyelid closure in some cases of facial nerve palsy, these patients should be evaluated for corneal injury. Patients with reduced eyelid closure and at risk for corneal abrasion should be considered for insertion of a gold weight to achieve eye closure. In patients where the nerve has been transected or resected, reconstruction of the facial nerve at that time should be considered when possible. If the proximal and distal nerve stumps are available, the nerve may be repaired with direct nerve coaptation at the time of tumor extirpation. If the repair cannot be achieved without tension, then an interpositional cable nerve graft or nerve transfer should be utilized. For the donor nerves, small diameter sensory nerves are commonly used, including the medial antebrachial cutaneous nerve, the sural nerve, and in some cases of facial nerve reconstructions, the greater auricular nerve (49–51). The donor nerves selected with depend on the length of the nerve deficit, the number of cables required, and the sensory deficit at the graft site. In some cases, the proximal or distal stump of the nerve may not be available and therefore nerve repair or graft may not be possible. However, nerve transfers such as the partial hypoglossal (52,53) to facial nerve or muscle transposition of the temporalis or masseter muscle are excellent alternatives and may be performed primarily (51,54–56). While controversy remains regarding postoperative radiation and nerve regeneration, several studies have documented no adverse effects (57–59). Gullane et al. (59) reported excellent function in four of six patients following facial nerve grafts and postoperative radiation after extensive resection of parotid cancers. It is our recommendation that patients who may undergo postoperative radiation should not be denied repair of the facial nerve. When facial nerve resection extends intratemporally or intracranially, it may not be possible to restore nerve continuity and other reconstructive options for facial reanimation should be considered. Delayed reconstruction may include a
145
free muscle transfer innervated by the trigeminal nerve motor branch to the masseter muscle, a cross facial nerve graft with a secondary free muscle transfer or muscle transposition of the temporalis or masseter muscle (51,54,55,60). Procedures that utilize a free tissue muscle transfer and do not involve reinnervation of the facial muscles may be performed even after long periods of facial nerve injury (51,54,55,60). In patients with a poor survival prognosis or who choose not to undergo a complex reconstruction, static facial slings may be a good alternative to restore facial symmetry at rest.
Cerebrospinal Fluid Leak A cerebrospinal fluid leak can have significant consequences with regard to flap compromise, flap loss, and patient morbidity. With increased exposure of the dura, there is an increased risk of a postoperative cerebrospinal fluid leak. Therefore, to minimize the risk of this complication, it is important to attain a watertight dural seal and this is best accomplished with the use of vascularized tissue for the reconstruction and meticulous closure of the repair.
Intracranial Infections The incidence of intracranial infections has decreased with the introduction of free flaps. However, if an intracranial infection does occur, it is important that broad-spectrum antibiotics are instituted promptly.
Late Complications Late complications are not life threatening and therefore are often underestimated in terms of patient morbidity. These complications include diplopia, malocclusion, trismus, nasal obstruction, and facial abnormality and often result from soft tissue atrophy, radiation fibrosis, and/or bony support. Correction of these deformities may be helpful to decrease patient morbidity and to improve health-related quality of life.
CONCLUSION Local and regional flaps have been used in the reconstruction of the skull base but these flaps are less frequently used for large defects. Because of the advantages of the free tissue transfer, our reconstructive method of choice for these defects in the skull base region is a free flap. The increased use of vascularized free flaps in these reconstructions has provided a significant decrease in postoperative complications and improved patient outcome (34,37). Skull base neoplasms often result in extensive ablation, which often produces large soft tissue defects with dural exposure and this will increase the risk of postoperative complications. A successful reconstruction in the skull base requires a watertight dural seal and obliteration of all dead space. This will effectively separate the intracranial contents from the upper aerodigestive tract and all potential infective sources. The use of a well-vascularized tissue will provide an excellent dural barrier and seal, in addition to obliterating potential sources of infection. At one time, skull base tumor ablation and reconstruction were considered too risky to even attempt but now this surgery is routine, reliable, and predictable in many patients. Advancements in diagnosis, imaging, and surgical technique, both for tumor resection and for reconstruction, have made this method of management achievable. The location of the tumor and the size of the skull base defect are two of the factors that will help to determine the optimal reconstruction. However, the use of a vascularized free tissue transfer for the reconstruction will help to achieve a watertight dural seal, obliterate any communication between the cranial contents and the aerodigestive tract, and thus
146
Neligan et al.
minimize the occurrence of postoperative complications in these patients. REFERENCES 1. Crile G. Excision of cancer or the head and neck with special reference to the plan of dissection based on one hundred and thirty two operations. JAMA. 1906;47:1780–1798. 2. Hobar PC, Barton FE. Head and neck II: Reconstruction. Selected Readings in Plastic Surgery. 1990: 6. 3. Ketcham AS, Hoye RC, Van Buren JM, et al. Complications of intracranial facial resections for tumors of the paranasal sinuses. Am J Surg. 1966;112:591–596. 4. Ward GE, Loch WE, Lawrence W. Radical operation for carcinoma of the external auditory canal and middle ear. Am J Surg. 1951;82:169–178. 5. Parson H, Lewis JS. Subtotal resection of the temporal bone for cancer of the ear. Cancer. 1954;7:995–1001. 6. Murray JE, Matson DD, Habal MB, et al. Regional cranio-orbital resection for recurrent tumors with delayed reconstruction. Surg Gynecol Obstet. 1972;134:437–447. 7. Golovine SS. Procede de cloture plastique de l’orbite apres l’exenteration. Arch d’Opthalmol. 1898;18:679–682. 8. Gillies HD. Plastic Surgery of the Face. London: Oxford University Press, 1920. 9. Ariyan S, Stahl RS. Reconstruction following cranio-orbital resection. In: Ariyan S, ed. Cancer of the Hand and Neck. St Louis, MO: C. V. Mosby Company, 1987:379–387. 10. Bakamjian VY, Souther SG. Use of the temporalis muscle flap for reconstruction after orbito-maxillary resections for cancer. Plast Reconstr Surg. 1975;56:171–177. 11. Holmes AD, Marshall KA. Uses of the temporalis muslce flap in blanking out orbits. Plast Reconstr Surg. 1979;63:336–343. 12. Sypert GW, Habal MB. Combined cranio-orbital surgery for extensive malignant neoplasms of the orbit. Neurosurgery. 1978;2:8–14. 13. Thomson HG. Reconstruction of the orbit after radial exenteration. Plast Reconstr Surg. 1970;45:119–123. 14. McGregor IA. The temporal flap in introral cancer: its use in repairing the post-excisional defect. Br J Plast Surg. 2004;16:318– 322. 15. Bakamjian VY. A two stage method for pharyngoesophageal reconstruction with a primary pectoral flap. Plast Reconstr Surg. 1965;36:173–184. 16. McGregor IA, Jackson IT. The extended role of the deltopectoral flap. Br J Plast Surg. 1970;23:173–185. 17. Jackson IT, Hide TAH. A systematic approach to tumors of the base of the skull. J Maxillofac Surg. 1982;10:92–98. 18. Lewis MB, Remensnyder JP. Forehead flap for reconstruction after ablative surgery for oral and oropharyngeal malignancy. Plast Reconstr Surg. 1978;62:59–65. 19. Park JS, Sako K, Marchetta FC. Reconstructive experience with the medially based deltopectoral flap. Am J Surg. 1974;128:548– 552. 20. Ariyan S. The pectoralis major myocutaneous flap. A versatile flap for reconstruction of the head and neck. Plast Reconstr Surg. 1979;63:73–81. 21. Ariyan S. Further experiences with the pectoralis major myocutaneous flap for the immediate repair of defects from excisions of head and neck cancers. Plast Reconstr Surg. 1979;64:605–612. 22. Quillen CG, Shearin JC Jr, Georgiade NG. Use of the latissimus dorsi myocutaneous island flap for reconstruction in the head and neck area. Case report. Plast Reconstr Surg. 1978;62:113–117. 23. Neligan PC, Boyd JB. Reconstruction of the cranial base defect. Clin Plast Surg. 1995;22:71–77. 24. Irish J, Gullane PJ, Gentili F, et al. Tumors of the skull base: Outcome and survival analysis of 77 cases. Head Neck. 1994;16:3– 10. 25. Jones NF, Schramm VL, Sekhar LN. Reconstruction of the cranial base following tumour resection. Br J Plast Surg. 1987;40:155– 162.
26. Sandor GK, Charles DA, Lawson VG, et al. Transoral approach to the nasopharynx and clivus using the Le Fort I osteotomy with midpalatal split. Int J Oral Maxillofac Surg. 1990;19:352– 355. 27. Ousterhout DK, Tessier P. Closure of large cribriform defects with a forehead flap. J Maxillofac Surg. 1981;9:7–9. 28. Westbury G, Wilson JSP, Richardson A. Combined craniofacial resection for malignant disease. Am J Surg. 1975;130:463–469. 29. Jackson IT, Marsh WR, Hide TAH. Treatment of tumors involving the anterior cranial fossa. Head Neck. 1984;6:901–913. 30. Scher RL, Cantrell RW. Anterior skull base reconstruction with the pericranial flap after craniofacial reconstruction. Ear Nose Throat J. 1992;71:210–212. 31. Noone MC, Osguthorpe JD, Patel S. Pericranial flap for closure of paramedian anterior skull base defects. Otolaryngol Head Neck Surg. 2002;127:494–500. 32. Jackson IT, Adham MN, Marsh WR. Use of the galeal frontalis myofascial flap in craniofacial surgery. Plast Reconstr Surg. 1986;77:905–910. 33. Schramm VL Jr, Myers EN, Maroon JC. Anterior skull base surgery for benign and malignant disease. Laryngoscope. 1979;89:1077–1091. 34. Neligan PC, Mulholland RS, Irish J, et al. Flap selection in cranial base reconstruction. Plast Reconstr Surg. 1996;98:1159–1166. 35. Rosen HM. The extended trapezius musculocutaneous flap for cranio-orbital facial reconstruction. Plast Reconstr Surg. 1985;75:318–327. 36. Califano J, Cordeiro PG, Disa JJ, et al. Anterior cranial base reconstruction using free tissue transfer: Changing trends. Head Neck. 2003;25:89–96. 37. Mulholland RS, Boyd JB, Irish J, et al. Flap selection in cranial base reconstruction—local, pedicled or free? Plast Surg Forum. 1993;62:265–266. 38. Teknos TN, Smith JC, Day TA, et al. Microvascular free tissue transfer in reconstructing skull base defects: Lessons learned. Laryngoscope. 2002; 112:1871–1876. 39. Urken ML, Turk JB, Weinberg H, et al. The rectus abdominus free flap in head and neck reconstruction. Arch Otolaryngol Head Neck Surg. 1991;117:857–866. 40. Sekhar LN, Schramm VL, Jones NF. Combined resection of large neoplasms involving the lateral and posterior cranial base. In: Sekhar LN, Schramm VL, eds. Tumors of the Cranial Base: Diagnosis and Treatment. New York: Futura Publishers, 1986. 41. Fisch U, Fagan P, Valavanis A. The infratemporal fossa approach for the lateral skull base. Otolaryng Clin North Am. 1984;17:513– 552. 42. Bakamjian VY, Long M, Rigg B. Experience with the medially based deltopectoral flap in reconstructive surgery of the head and neck. Br J Plast Surg 1971;24:174–183. 43. Patel SG, Singh B, Polluri A, et al. Craniofacial surgery for malignant skull base tumors. Cancer. 2003;98:1179–1187. 44. Nahabedian MY, Singh N, Deune EG, et al. Recipient vessel analysis for microvascular reconstruction of the head and neck. Ann Plast Surg. 2004;52:148–155. 45. Lipa JE, Butler CE. Enhancing the outcome of free latissimus dorsi muscle flap reconstruction of scalp defects. Head Neck. 2004;26:46–53. 46. Badie B, Preston JK, Hartig GK. Use of titanium mesh for reconstruction of large anterior cranial base defects. J Neurosurg. 2000;93:711–714. 47. Imola MJ, Sciarretta V, Schramm VL. Skull base reconstruction. Curr Opin Otolaryngol Head Neck Surg. 2003;11:282–290. 48. Seddon HJ. Three types of nerve injury. Brain. 1943;66:237–238. 49. Sunderland S. Nerve and Nerve Injuries. Edinburgh: Churchill Livingstone, 1978. 50. Mackinnon SE, Dellon AL. Surgery of the Peripheral Nerve. New York: Thieme Medical Publishers, 1988. 51. Myckatyn TM, Mackinnon SE. The surgical management of facial nerve injury. Clin Plast Surg. 2003;30:307–318. 52. Kalantarian B, Rice DC, Tiango DA, et al. Gains and losses of the XII-VII component of the “babysitter” procedure: a morphometric analysis. J Reconstr Microsurg. 1998;14:459–471.
Chapter 8: Reconstruction of Skull Base Defects 53. Mersa B, Tiango DA, Terzis JK. Efficacy of the “babysitter” procedure after prolonged denervation. J Reconstr Microsurg. 2000;16:27–35. 54. Conley J, Gullane PJ. Facial rehabilitation with termporal muscle. New concepts. Arch Otolaryngol. 1978;104:423– 426. 55. Gullane PJ. Extratemporal facial rehabilitation. J Otolaryngol. 1979;8:477–486. 56. Manktelow RT, Tomat LR, Zuker RM, et al. Smile reconstruction in adults with free muscle transfer innervated by the masseter motor nerve: Effectiveness and cerebral adaptation. Plast Reconstr Surg. 2006;118:885–899.
147
57. Brandt K, Evans GRD, Ang KK, et al. Postoperative irradiation: Are there long-term effects on nerve regeneration ? J Reconstr Microsurg. 1999;15:421–425. 58. Evans GRD, Brandt K. Peripheral nerve regeneration: The effects of postoperative irradiation. Plast Reconstr Surg. 2003;112:2023– 2024. 59. Gullane PJ, Havas TJ. Facial nerve grafts: Effects of postoperative irradiation. J Otolaryngol. 1987;16:112–115. 60. Manktelow RT. Free muscle transplantation for facial paralysis. Clin Plast Surg. 1984;11:215–220. 61. Gullane PJ, Lipa JE, Novak CB, et al. Reconstruction of skull base defects. Clin Plast Surg. 2005;32(3):391–400.
9 Prosthetic Rehabilitation of Patients Undergoing Skull Base Surgery Theresa M. Hofstede, Rhonda F. Jacob, Pattii C. Montgomery, Peggy J. Wesley, Jack W. Martin, and Mark S. Chambers
alveolar ridge, and hard palate as supporting tissues for prostheses. The important supporting mandibular landmarks are the alveolar ridge, retromolar pad, and buccal shelf. Preserving, enhancing, or reconstructing these tissues is important for support and retention of a prosthesis. Periodontally healthy teeth and edentulous ridges are essential structures for prosthesis retention and support. Conservation of the supporting tissues, if possible, after eliminating disease, is important for optimal outcome; albeit, treatment of disease takes priority (9–11).
This chapter presents current concepts regarding oral and facial prosthetic rehabilitation and oral oncologic principles in patients with head and neck cancer. Multidisciplinary therapeutic techniques are commonly used in the care of patients with advanced disease of the head and neck. The surgical technique is used to eliminate local disease and aid in the preservation of function, mobility, and physical appearance. Hence, collaboration and team efforts between the head and neck surgeon, plastic reconstructive surgeon, maxillofacial prosthodontist, speech pathologists, and other vital members are important in improving a patient’s quality of life. Prior to the surgical procedure, the head and neck team specialists formulate plans and preparations for prosthetic rehabilitation. This can help prevent and minimize postsurgical complications and reconstructive dilemmas. Clear and open communication between the head and neck surgeon and the maxillofacial prosthodontist is of utmost importance for successful prosthetic rehabilitation (1–5). Early dental intervention can help reduce future challenges such as oral infection by decreasing risk factors for oral complications, such as periodontal disease, osteoradionecrosis, and caries. For this reason, patients are referred to the maxillofacial prosthodontist as early as possible for evaluation of their oral and dental condition and for discussion of treatment options in prosthetic rehabilitation (5,7). In some cases, for instance, ablated structures can be replaced immediately with prostheses that restore function and improve appearance. The primary head and neck surgeon integrates this information into the treatment plan. Essentially, rehabilitation of function and improved physical appearance as well as reduction of posttreatment sequelae are the primary aims after elimination of disease (1,3).
ORAL AND DENTAL EVALUATION A preoperative evaluation should include medical history, intercurrent illnesses, oral and dental status, previous chemotherapy or radiation treatment, and other factors, such as age, nutritional status, and history of tobacco and alcohol use (7,11,12). The oral and dental evaluation is part of the routine head and neck examination. The head and neck surgeon will identify acute or chronic pathologic conditions related to the dentition or supporting structures, such as advanced periodontal disease, gross dental caries, tissue irritation from poorly fitting prostheses, and poor oral hygiene (1). Gross caries, plaque, and calculus formation on teeth indicate poor oral hygiene, possibly periodontal disease, and a precipitating factor for sepsis. Oral and dental abnormalities are documented during the initial medical examination (7). The patient is then referred to a maxillofacial prosthodontist for further evaluation and treatment. The initial oral and dental examination by the prosthodontist confirms the existence of acute and chronic oral pathologic conditions, such as dental abscess, teeth with advanced periodontal disease, or dental calculus causing gingivitis (1,7). Diagnostic radiographs are required to evaluate any tumor involvement in the bone, periodontal bone loss, periapical or periodontal abcesses, and gross carious lesions. The more common diagnostic imaging used in an oral and dental examination are the panoramic, periapical, bitewing, and occlusal radiography. A panoramic radiograph shows the overall topographic features of the dentition, maxilla, mandible, sinuses, nasal cavity, and temporomandibular joints (1,13,14). These radiographs can be of great value to the head and neck surgeon in diagnosis of bony invasion by tumor of the maxilla or mandible and are easily obtained from most dental offices. Periapical, bitewing, and occlusal radiographs show a more detailed view of the teeth and bone. Stone casts obtained from impressions of the maxilla and mandible during the initial dental visit can be useful if a surgical prosthesis (intraoral, extraoral, or both) is needed, such as a surgical obturator for a patient who has to undergo
ORAL AND DENTAL ANATOMY Communicating the patient’s oral and dental status is of paramount importance during treatment planning between the surgeon and maxillofacial prosthodontist. It is important that the surgeon knows which teeth are to be extracted for the surgical procedure so that the maxillofacial prosthodontist can plan and fabricate an immediate or postoperative prosthesis. The universal numbering system of teeth is used for this purpose (Fig. 1). In this system, adult teeth are numbered sequentially from 1 to 32, starting with the right maxillary third molar (#1), going to the left maxillary third molar (#16), continuing with the opposing left mandibular third molar (#17), and finishing with the right mandibular third molar (#32) (8). Anatomic landmarks important in prosthetic rehabilitation in the edentulous maxillary arch are the tuberosity, 149
150
Hofstede et al.
Figure 1 The universal numbering system of teeth.
maxillectomy (15–17). Teeth having a poor or uncertain prognosis are identified and extracted before or during the primary ablative surgical procedure (17).
PREPROSTHETIC PRINCIPLES Radiation therapy to the head and neck often causes collateral damage to the salivary glands causing reduced salivary function in varying degrees and increased susceptibility to dental caries and oral infection (1–3). The extent of the morbidity is related to radiation dose, volume of tissue treated, and age of the patient when treated (1,7). These patients need to maintain meticulous oral hygiene and be placed on a caries preventive regimen (i.e., daily fluoride) and sialogogues for the rest of their lives. Numerous studies have shown that fluoride reduces radiation decay of teeth if used in a systematic and predictable manner (1,16,18). Caries are an important complication for patients with autoimmune- or radiation-induced xerostomia (1,18). Reducing the risk of dental infection while maintaining optimal oral health can decrease the risk of osteoradionecrosis or a septic condition in the care of a patient who undergoes radiation therapy or chemotherapy (3,6,19–24). Some patients who have ablative surgery due to head and neck cancer need removable (maxillofacial) prostheses to replace anatomic structures (1,2,16). There are usually three phases of prosthetic rehabilitation: surgical, interim, and definitive. Each phase includes fabrication or modification of a prosthesis. These phases span several months to a year. Surgical and interim prostheses may have to be adjusted frequently while tissues are healing (1,15). In the maxillectomy patient, the surgical obturation prostheses are used to restore oral contour and function immediately after maxillectomy. The obturator separates the oral and nasal cavities to allow normal speech and swallowing. A common problem associated with maxillary obturator prostheses is leakage of fluid around the prosthesis and through the nose. Speech and swallowing can be compromised but in most cases can be corrected immediately during the postoperative period.
Most patients are hesitant about starting oral and dental hygiene postoperatively for fear of disturbing the surgical site. However, meticulous oral, dental, and prosthetic hygiene is encouraged immediately after head and neck surgery to reduce oral microbes and associated infections (1,25). Routine oral and dental hygiene can be initiated 2 weeks after surgery; however, initial postoperative oral care should be limited to oral lavage and decontaminating rinses (1,6,25). Postoperative physiotherapy for trismus is a consideration in rehabilitation. Suitable techniques can be discussed and explained to the patient before surgery and reinforced postoperatively. Such therapy can maintain the oral opening and allow the patient better access to the surgical defect and to the rest of the oral cavity (1,25,26). Opening exercises with wooden tongue blades or sophisticated oral opening devices R (i.e., TheraBite oral opening device, Atos Medical Co., R Milwaukee, WI; Dynasplint , Dynasplint Systems, Inc., Severna Park, MD) can be effective in maintaining or restoring oral opening following surgery (1–3,17). Reduced mobility of the jaw is a serious and sometimes painful condition that prohibits patients from doing simple tasks such as chewing, swallowing, speaking, or maintaining oral hygiene. Pain can be persistent and coupled with limited functioning. The patient’s quality of life is significantly affected. Oral opening exercises can improve jaw opening by stretching connective tissues, mobilizing joints, strengthening muscles, and activating anti-inflammatory properties, leading to reduction of pain and inflammation. A physiotherapist may be consulted in more complex cases to implement electrotherapy, lymphedema therapy, ultrasound, and other advanced techniques as needed (1–3,17). Complications vary and depend on the patient’s oral and dental status, type of malignancy, and type of oncologic therapy. Thorough oral and dental assessment and treatment by a maxillofacial prosthodontist can greatly minimize oral complications, promote oral health, and optimize the reconstruction outcome.
SURGICAL TECHNIQUES TO ENHANCE PROSTHETIC REHABILITATION Maxilla Resected lesions of the maxillary sinus, hard palate, and alveolar ridge can cause postoperative speech and swallowing problems. These difficulties can be minimized or eliminated with careful surgical planning (1–3). The dental cast obtained at the initial dental visit is used in discussion with the surgeon for the surgical planning and fabrication of the surgical prosthesis. Several surgical techniques can be incorporated into maxillectomy procedure to improve prosthetic rehabilitation. 1. Alveolar osteotomies are made through the socket of an extracted tooth or an edentulous space to prevent iatrogenic bone loss and to ensure the longevity of the tooth next to this osteotomy (1,15,16). 2. When making the palatal osteotomy, as much of the premaxilla as possible is spared. The premaxilla is important for the support and retention of a prosthesis. If the tumor is in the anterior region of the sinus, it may be possible to spare the maxillary tuberosity on the side of the defect, which would also increase prosthetic support. 3. A split-thickness skin graft or allogenic material is placed in the maxillary defect (1,27). The skin graft provides an excellent scar tissue band for retention of the prosthesis, and decreases mucus secretion, nasal polyps, and crust
Chapter 9: Prosthetic Rehabilitation of Patients Undergoing Skull Base Surgery
4. 5.
6.
7.
8.
formation in the ablated sinus; the result is improved hygiene. The palatal mucosa, if not affected by disease, can be retained and wrapped around the midline portion of the palatal cut. Removal of the inferior and middle turbinates allows proper extension of the prosthesis into the defect area. If not removed, the turbinates can be irritated by the prosthesis, and be a constant source of irritation and bleeding (1–3,17). Consideration is given to removing the posterior mandibular molars on the side of maxillectomy. These teeth can cause a hygiene problem and are essentially nonfunctional after maxillectomy (25). The Weber–Fergusson incision is used in some cases for access in a maxillectomy; however, an intraoral approach is increasingly becoming utilized in eliminating the facial incision. This makes manipulation of the lip and cheek easier for the dentist and patient during postoperative prosthetic procedures (1,9,15,17,27). A surgical obturator prosthesis is placed to restore the oral contour for immediate function and good postoperative appearance. This prosthesis supports the surgical packing and can be fixed to the remaining teeth with surgical wire or retained with a bone screw in an edentulous patient. Use of an obturator can obviate use of a nasogastric tube and decrease the duration of postoperative rehabilitation. The obturator maintains proper lip and cheek support during healing and helps to reduce contracture of scar tissue. When the surgical packing is removed, usually within 5 to 7 days of the procedure, the surgical obturator can be converted into an interim prosthesis. Use of a surgical prosthesis can improve the patient’s psychologic status, as it allows them to speak and eat following surgery (1,9,15,27).
As previously discussed, there are three distinct phases of maxillary obturator rehabilitation: surgical, interim, and definitive (1). The size and location of the surgical defect, dentition status, and supporting surface area of remaining palate and overlying structures determine the stability and retention of an obturator (1,2,6,10,17,28). As the size of the defect increases and the residual palatal tissues decrease, stability and functionality of the prosthesis reduce significantly due to the lack of teeth for direct clasp retention or surface area to engage undercuts (1,2,6,10,28). Some clinicians suggest that maxillary defects should be closed completely using free-tissue grafts (1,6,17,28,29). Function and cosmesis can be effectively restored with surgical reconstruction. This method does occlude the surgical defect but may preclude prosthetic rehabilitation and may become a challenge if fistulation develops or flap bulkiness occurs, thus, disallowing oral function and prosthetic placement. Surgical reconstruction is an excellent alternative for patients if a microvascular osteocutaneous flap is harvested and implants can be placed to retain a prosthesis (1,6,29–31).
SURGICAL RECONSTRUCTION WITH MICROVASCULAR TECHNIQUES Maxillary reconstruction provides a separation between the oral cavity from the sinus and nasal cavities, thus eliminating problems associated with obturator prostheses. Current microvascular techniques have facilitated the use of osteocutaneous free flaps to reconstruct orofacial defects by providing adequate amounts of bone and soft tissue in a single
151
surgical procedure (32). Since the use of the first fibula free flap, surgical reconstruction of orofacial defects has become a routine procedure (33,34). Numerous cases of mandibular reconstruction have been reported, whereas accounts of bilateral maxillary reconstruction are more rare (34–39). Successful reconstruction of the maxilla requires an osseous element that relays occlusal forces to the cranium, resists long-term resorption, and allows esthetic and functional rehabilitation with implants [Fig. 2(A)–2(D)] (31,40). The cutaneous element of the reconstruction should separate the oral cavity from the sinus/nasal cavities, thereby allowing proper speech and deglutition (31,40). The free fibula flap is well suited for reconstruction of the both the mandible and the maxilla. The fibula provides an adequate length of bone that can be segmented to replace the contour of the alveolar ridge. The length is sufficient for reconstruction of large defects. The bicortical configuration and vertical height of the fibula are ideal for endosseous implant placement. Studies have found the vertical height of the fibula to be between 13.1 and 16.7 mm, with an average of 15.2 mm (31,40,41). This can provide bicortical fixation of implants that can ultimately improve osseointegration. Long-term observation has found only 2% to 7% bone resorption in the vascularized fibula (42). The soft pliable skin paddle can be used for reconstruction of intraoral or cutaneous tissue (32).
Soft Palate When some soft palate is removed in the surgical procedure, the surgeon must consider whether the remaining soft palate will be functional. It is easier to rehabilitate a patient’s speech and swallowing if the soft palate is removed totally (1,6). If the remaining soft palate is nonfunctional, rehabilitation can be difficult or even impossible. Sometimes a thin strip of soft palate can be useful for prosthesis retention in a patient with limited supporting tissue (1, 2,9,17). Primary irradiation of the soft palate can cause palatal incompetency due to fibrosis and tumor necrosis. Patients with this dilemma may regurgitate liquid and food through the nose, and speech may be hypernasal (6,9,17). In this situation, prosthetic rehabilitation can be impossible because of poor access to the oral pharynx.
Extraoral Lesions Lesions involving facial structures can necessitate prosthetic rehabilitation. The reconstructive surgeon must understand both the limitations of facial prostheses and the surgical techniques that enhance the adaptation and appearance for successful prosthetic rehabilitation. The restoration of facial defects is a difficult challenge. As previosuly stated, the success of a maxillofacial or facial prosthesis can be greatly enhanced by careful presurgical evaluation and communication involving the patient, surgeon, and maxillofacial prosthodontist (43,44). Facial prostheses can be made from a variety of materials, such as polymethyl methacrylate or medical-grade platinum silicone elastomer with or without a polyurethane backing (45). These prostheses are retained with medical-grade prosthetic adhesives, medical-grade double-sided tape, anatomic undercuts, acrylic substructures with magnets, or in some cases, extraoral osseointegrated implants or a combination thereof (1,43,46). Facial and intraoral prostheses can be connected with magnets to aid in retention and functionality. The esthetic result depends on the amount of tissue removed, type of reconstruction, morbidity of adjunctive treatment (radiation and chemotherapy), and the physical characteristics of the tissue base available to support and retain the prostheses (1,17,47).
152
Hofstede et al.
(A)
(B)
(C)
(D)
Figure 2 (A) A 15-year-old patient with history of osteosarcoma of left maxilla underwent infrastructure maxillectomy and delayed reconstruction with a free fibula microvascular flap. (B) Implants were placed 3 months following fibula reconstruction. (C) Transmucosal implant attachments ready for prosthesis impression. Granulation tissue through the skin flap is common and may require limited surgical removal, repeated application of topical steroids, and aggressive hygiene practices. (D) Transitional obturator prosthesis supported by dentition and implants.
Orbit In total orbital exenteration, several surgical considerations can improve prosthetic rehabilitation [Fig. 3(A) and 3(B)]. The eyelid is resected while the position of the eyebrow is
maintained. Sharp or rough bony margins are smoothed and rounded. If possible, the infraorbital bony margin is reconstructed. A split-thickness skin graft is placed into the area of the defect to cover exposed bone creating a concavity to house the prosthesis. Sufficient depth of defect is essential in order to fabricate an esthetic orbital prosthesis. Hygiene becomes easier for the patient, and the prosthesis can be extended into the defect for greater orientation and stability (1–3,46).
Nose
(A)
(B)
Figure 3 (A) A 64-year-old man with a history of neuroendocrine small cell carcinoma of the right orbital region; treated with orbital exenteration, reconstruction with a split-thickness skin graft, and postoperative external beam radiation treatment. (B) Final elastomeric orbital prosthesis is supported by a prosthetic substructure within the orbital defect.
Resection of lesions involving the nose can necessitate partial or total nasal restoration [Fig. 4(A) and 4(B)]. The nasal spine is left intact, if possible, for stability of the nasal prosthesis. Unsupported tissue tags are removed, because they can make impression techniques difficult and can compromise the final prosthesis by hindering the patient in placing and adhering a prosthesis. Rough tissue margins or distortion of adjacent facial contours compromises the concealment of the prosthetic margins and retention. A split-thickness skin graft is placed over the resected bone margins to increase the stability of the nasal prosthesis (1–3). Grafts or flaps are used to maintain the position of the upper lip.
Ear Operations on the ear can vary from subtotal resection to total auriculectomy [Fig. 5(A)–5(C)]. It is easier to replace a complete ear than a partial ear with a prosthesis. With a
Chapter 9: Prosthetic Rehabilitation of Patients Undergoing Skull Base Surgery
153
troublesome when the patient attempts to insert or place the prosthesis (43). The superior half of the auricle has better cartilaginous support yet tends to become distorted after the operation. This distortion is accentuated when the residual auricle is rotated and used to close the defect. A preserved portion of the root of the helix is a good landmark and support for eyeglasses (1). This area can help later in vertical support of the prosthesis. The anterosuperior helical rim is left in place if possible. Posterior regions can be grafted. Though surgical techniques can enhance the recipient tissue bed for the prostheses, it is understood that the main concern of the treating surgeon is elimination of disease.
(A)
(B)
Figure 4 (A) A 82-year-old man with a history of invasive squamous cell carcinoma of the nose, status post complete rhinectomy with a partialthickness skin flap reconstruction. (B) Final nasal prosthesis, supported by medical grade adhesives, provides form and function.
total replacement, the maxillofacial prosthodontist has more freedom in shape, size, and location of the prosthesis. First, the recipient area must be flat or concave. Convexities from excessive tissue bulk can hamper esthetic results. Second, skin devoid of hair provides a good adhesive base, although a split-thickness skin graft is better. Tissue pockets assist in orientation and stabilization of the prosthesis and allow the margins to extend in a 0-degree emergence profile (1–3,43). If tissue can be spared, the tragus is the first choice. It is a good separate landmark that is not easily displaced (1–3,43). The tragus allows the anterior margin of the prosthesis to be hidden behind the posterior flexure. It also aids the patient in proper positioning of the prosthesis by providing a placement reference. Hair, the angle of the helical rim, or both provide posterior margin concealment. The inferior half of the soft-tissue pinna is of little or no use. Because it lacks cartilaginous support, the lobe of the auricle is normally drawn down and away from the head. It is difficult to capture this effect in an impression, and bilateral symmetry usually cannot be achieved. The lobe margin is difficult to maintain and can be
(A)
(B)
OSSEOINTEGRATED DENTAL IMPLANTS Osseointegrated dental implants have become an excellent alternative to conventional prosthetic rehabilitation (1,17,31,48–50). The principle of osseointegration of commercially pure titanium through the intermediary of the layer of titanium oxide and integration with surrounding bone opened a completely new field of clinical application for biotechnology in prosthetic dentistry [Fig. 6(A)–6(D)] (51,52). Implant designs fall into three main categories: endosseous, subperiosteal, and transosteal. There are over 50 implant subtypes from which the prosthodontist can choose in rehabilitating the patient (49,51). Additionally, there are numerous abutments available that range from fixed (cemented) and fixed-removable (screw or clip retained) designs to removable affixed by o-rings, clips, or snap designs (1,49). The most common implants used today are endosseous implants that are surgically inserted into the maxillae or mandible and integrate with the surrounding bone (50). Endosseous implants have provided the support, retention, and stability that is greatly needed in compromised oral cavities following tumor ablative procedures (1,17,31,50). Several common terms are used today in oral implantology: dental implant, endosteal dental implant, abutment, and osseointegration (1,51). A dental implant is a device, specially designed to place surgically within or on the mandibular or maxillary bone in providing resistance to displacement of an overlying prosthesis. It can be placed transgingivally or fully embedded subgingivally (1,51). The
(C)
Figure 5 (A) A 72-year-old man with a history of basal cell carcinoma of the right ear, status post subtotal auriculectomy, reconstructed with a full-thickness skin graft. (B) Final auricular prosthesis restores esthetic appearance. (C) The auricular prosthesis has an intimate fit with remaining superior tissue tag.
154
Hofstede et al.
(A)
(B)
(C)
(D)
Figure 6 (A) Patient with a previous iliac crest flap mandibular reconstruction and endosteal implants without benefit of medical models or 3-D imaging. (B) Twenty years later, the patient has a second primary in the floor of mouth and requires resection of the iliac crest flap and another microvascular reconstruction. A 3-D model is created, surgery is performed, and a mandibular template is created for preforming a reconstruction plate. (C) Panoramic radiograph reveals six osteotomies in the microvascular fibula flap to gain appropriate mandibular adaptation to the reconstruction plate. (D) Mandibular contour without benefit of 3-D modeling (left) and second mandibular reconstruction 20 years later with 3-D modeling and model surgery (right).
endosteal dental implant is an implant of which a part is anchored within the maxillary or mandibular bone (17,31,51). An abutment is a component connected to the endosteal part of the dental implant and allows retention or support of a dental implant (17,31,51). The abutment connects the fixture (implant) to a superstructure and overlying prosthesis (49–51). Osseointegration is a direct structural and functional connection between vital bone and the surface of a load-carrying implant (51, 52). Finally, the endosseous dental implants can be divided into different categories according to their shape, surface characteristics, chemical composition, or the way the implant is inserted into the bone (1,51). Meticulous, surgical, and prosthetic planning is required prior to the placement of implants (50). If adequate supporting tissues exist or can be established, patients may function satisfactorily with routine prostheses, or they may be considered for implants if the recipient sites are appropriate. There are several types of implants currently utilized for intraoral and extraoral maxillofacial uses (1,31,47). Most implants used in maxillofacial prosthetics are made of titanium and are cylindrically designed utilizing threads that anchor them in the bone. Such implants are produced in a variety of lengths and widths. In most of these implant systems, placement involves a two-stage procedure. The first
stage, performed under local or general anesthesia, is the placement of the implant into the bony recipient site (1,49,50). Placement is performed in a very precise fashion to ensure the least amount of damage to the adjacent bone and soft tissue, particularly in irradiated tissues. Following placement, the implant is covered primarily by the initial tissue flap and allowed to integrate with the bone for 12 or more weeks. The purpose is to promote peri-implant integration with the implant interface as described previously. Confirmed osseointegration, by radiographic and clinical interpretation, implies a direct and lasting connection between vital bone and the titanium implant of defined surface topography and geometry (1–3). In the second stage, only the superior portion of the implant is uncovered and an abutment, usually made of titanium alloy, is placed onto the implant (1). This connection will join the implant to the prosthesis. The number of implants and type of retention as well as prosthetic designing are at the discretion of the restoring prosthodontist. In general, the prosthesis is fabricated so that it can be easily removed by the patient and allow maintenance and proper oral hygiene. Most irradiated maxillofacial patients are restored using removable prostheses (1,47). For those patients, implant rehabilitation can be successfully restored with an overlying removable prosthesis using
Chapter 9: Prosthetic Rehabilitation of Patients Undergoing Skull Base Surgery
ball abutments and o-ring attachments embedded in the prosthesis. Implants can be placed during or after the primary reconstructive procedure if the host bone has acceptable cortical plates suitable for fixation of implants (53,54). The fibula tissue graft is an excellent recipient site for such implants (55). The fibula free flap provides a long segment of bone and can include a large fasciocutaneous component (55–60). As such, this versatile flap may be harvested as an osteomyocutaneous flap or a purely osseous flap. The pedicle runs the length of the fibula, with perforators extending to the skin paddle (1,55–60). It provides the longest segment of bone currently available for harvest; up to 26 cm can be taken without affecting leg function. Many bony defects of the head and neck region are well suited for reconstruction with a fibula free flap. Functional rehabilitation can be optimal. Patients who have osseointegrated implants placed in the fibula graft can have near-normal mastication and speech re-established (55,56). In most cases, the placement of implants is best accomplished during a secondary procedure when a premade surgical stent can guide the prosthodontist in aligning and positioning the implants (1). Patients who have received postoperative radiation in proximity to the proposed implant
(A)
(C)
155
site may be poor candidates due to hypovascularity in the volume of tissue radiated and may benefit from hyperbaric oxygen therapy prior to implant placement (19,61–66). The goals of prosthetic rehabilitation with implants include predictability and simplicity. Fabrication of a fixed implant prosthesis often involves methods and materials that are complicated and increase costs to the restoring dentist and patient. Additional prosthetic components as well as additional appointments to complete the prosthesis can create unnecessary inconveniences; therefore, a removable prosthesis should be considered initially in treatment planning for a patient with maxillofacial deficits (1,50). Aside from patient factors, decision-making on the type of implant to use as well as restoring materials and prosthetic rehabilitation should be based on evidence from clinical protocols and facts and figures of corporate products. Improvements in dental implants, biomaterials, bioengineering, and preserving-surgical techniques increase the retention, stability, and longevity of prosthetic rehabilitation and ultimately a patient’s quality of life (1,17,67). The use of osseointegrated implants and guided tissue regeneration (i.e., synthetic membranes) procedures in irradiated patients have great promise (48,68). The cellular and metabolic changes in
(B)
(D)
Figure 7 (A) Microvascular reconstruction with a fibula flap involving the entire hard palate in this irradiated patient who lost left maxilla due to adenoid cystic carcinoma and later lost right maxilla due to osteoradionecrosis. Fibula and skin paddle created the separation between the oral cavity and paranasal sinuses. (B) Reformatted 3-D images of microvascular fibula flap. (C) Panoramic radiograph reveals fibula flap and subsequent implant reconstruction. (D) Full face of patient who has undergone a reconstructed maxillae and dental rehabilitation.
156
Hofstede et al.
the irradiated bone may affect both the quality and quantity of osseointegration (1,47). Important factors in successful dental restoration are the influence of radiation dose, delivery system (e.g., intensity modulated radiation therapy, IMRT), implant biomaterials, time from radiation therapy to implant surgery, the experience of the prosthodontist, fixture and abutment lengths, appropriate implantation sites, and type of prosthetic retention on implant survival (1–3). In addition, the benefit of osteogenesis and angiogenesis of hyperbaric oxygen therapy in conjunction with reconstructive techniques in the irradiated patient is also being studied (61,66). New physiotherapy methods are being developed to improve the postoperative function of patients with head and neck cancer (1–3,26). Tissue engineering, stereolithography, and rapid prototyping technology have recently resulted in the successful formation of new tissue equivalents of bone and cartilage that will enhance prosthetic rehabilitation in the future (68–71). Stereolithography (SL) is a computerized technology that uses computed tomography (CT) scans to fabricate an accurate three-dimensional model of a structure. This technology has been used extensively to preplan craniofacial surgery (72). SL models are reported to have an accuracy of 97.9%, with an average difference of 0.12 mm on mandible specimens (73). A study of SL models made from cadaver skull specimens found that the mean overall difference ranged from 0.8 to 2.5 mm (74). The accuracy is improved with smaller cuts during CT acquisition. Such accuracy has allowed surgeons to create templates to harvest bone flaps and prefabricate plates for surgical reconstruction [Fig. 7(A)–7(D)] (75–77).
CONCLUSIONS Fostering communication and compliance among members of the multidisciplinary treatment team will ensure quality of preventive, therapeutic, and maintenance care to patients with head and neck cancer. Thoughtful collaboration among the surgical oncologist, radiation and medical oncologist, maxillofacial prosthodontist, and implant surgeon optimizes the delivery and timing of dental care; therefore, improving the quality of the prosthetic rehabilitation. A functional, esthetic dental prosthesis after cancer treatment is a critical component of patient management that focuses on improving quality of life. To achieve these goals, each member of the multidisciplinary team must have working knowledge of the principles of prosthetic rehabilitation and appreciate the potential adverse impact that a suboptimal prosthesis could have on the patient’s functional, esthetic, and emotional outcome. The dental status has a persistent impact on quality of life in patients who have been successfully treated for head and neck cancer. REFERENCES 1. Chambers MS, Lemon JC, Martin JW. Surgical techniques to enhance prosthetic rehabilitation. In: Bailey BJ, Johnson JT, Newlands S, eds. Head and Neck Surgery-Otolaryngology. Philadelphia, PA: Lippincott Williams and Wilkins, 4th edition, Chapter 127; 2006:1853–1865. 2. Lemon JC, Martin JW, Jacob RF. Prosthetic rehabilitation. In: Weber RS, Miller MJ, Goepfert H, eds. Basal and squamous cell skin cancers of the head and neck. Baltimore: Williams & Wilkins, 1996:305–312. 3. King GE, Jacob RF, Martin JW. Oral and dental rehabilitation. In: Johns ME, ed. Complications in otolaryngology head and neck surgery. Philadelphia, PA: BC Decker, 1986:131.
4. Van Blarcom CW. The glossary of prosthodontic terms, 8th ed. J Prosthet Dent. 2005;94:51. 5. Jacob RF. The traditional therapeutic paradigm: Complete denture therapy. J Prosthet Dent. 1998;79:6–13. 6. Jacob RF, King GE. Indirect retainers in soft palate obturator design. J Prosthet Dent. 1990;63:311–315. 7. Lingeman RE, Singer MJ. Evaluation of the patient with head and neck cancer. In: Sven JY, Myers EN, eds. Cancer of the head and neck. New York: Churchill Livingstone, 1981:15. 8. Fuller JL, Deneky GE. Concise dental anatomy and morphology, 2nd ed. Chicago, IL: Mosby Year Book, 1984:9. 9. Beumer JP, Curtis TA. Restoration of acquired hard palate defects. In: Maxillofacial rehabilitation: prosthodontic and surgical considerations. St. Louis: Mosby, 1979:188. 10. Okay DJ, Genden E, Buchbinder D, Urken M. Prosthodontic guidelines for surgical reconstruction of the maxilla: A classification system of defects. J Prosthet Dent. 2001;86:352– 363. 11. Genden EM, Okay D, Stepp MT, Rezaee RP, Mojica JS, Buchbinder D, Urken ML. Comparison of functional and qualityof-life outcomes in patients with and without palatomaxillary reconstruction. Arch Otolaryngol Head Neck Surg. 2003;129:775– 780. 12. Silverman S. Oral Cancer: Complications of therapy. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1999;88:122–126. 13. Langland OE, Langlois RP, Morris CR. Principle and practice of panoramic radiology. Philadelphia, PA: WB Saunders, 1982:131–156. 14. Blaschbe DP, Osborn AG. The mandible and teeth. In: Bergeron RT, Osborn AG, San PM, eds. Head and neck imaging. St. Louis: Mosby, 1984:279. 15. Martin JW, Jacob RF, Larson DL, et al. Surgical stents for the head and neck cancer patient. Head Neck Surg. 1984;7:44. 16. Curtis TA, Beumer J. Restoration of acquired hard palate defects: Etiology, disability, and rehabilitation. In: Beumer J, Curtis TA, Firtell DN, eds. Maxillofacial Rehabilitation: Prosthodontic and Surgical Considerations. St. Louis: CV Mosby Co., 1979:188– 243. 17. McCord JF, Michelinakis G. Systematic review of the evidence supporting intra-oral maxillofacial prosthodontic care. Eur J Prosthodont Restor Dent. 2004;12(3):129–135. 18. Fleming TJ. Oral tissue changes of radiation oncology and their management. Dent Clin North Am. 1990;34:233–237. 19. Marx RE, Johnson RP. Studies in the radiobiology of osteoradionecrosis and their clinical significance. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1987;64:379–390. 20. Marx RE. Radiation injury to tissue. In: Kindwall EP, ed. Hyperbaric medicine practice. Flagstaff, AR: Best Publishing, 1994;447–503. 21. Marx RE, Ehler WJ, Tayapongsak P, et al. Relationship of oxygen dose to angiogenesis induction in irradiated tissue. Am J Surg. 1990;160:519–524. 22. Marx RE, Johnson RP, Kline SN. Prevention of osteoradionecrosis: A randomized, prospective clinical trial of hyperbaric oxygen versus penicillin. J Am Dent Assoc. 1985;11:49–54. 23. Ang E, Black C, Irish J, et al. Reconstructive options in the treatment of osteoradionecrosis of the craniomaxillofacial skeleton. Br J Plas Surg. 2003;56:92–99. 24. Marx RE, Johnson RP. Problem wounds in oral and maxillofacial surgery: The role of hyperbaric oxygen. In: Davis JC, Hunt TK, eds. Problem Wounds: The Role of Oxygen. New York: Elsevier, 1988:65–123. 25. Martin JW, Austin JR, Chambers MS, et al. Postoperative care of the maxillectomy patient. ORL Head Neck Nurs. 1994;12:15–20. 26. Barrett NV, Martin JW, Jacob RF, et al. Physical therapy techniques in the treatment of the head and neck patient. J Prosthet Dent. 1988;59:343. 27. Teichgraeber J, Larson DL, Castaneda O, et al. Skin grafts in intraoral reconstruction: A new stenting method. Arch Otolaryngol Head Neck Surg. 1984;101:463. 28. Aramany MA. Basic principles of obturator design for partially edentulous patients. Part II: design principles. J Prosthet Dent. 1978;40:656–662.
Chapter 9: Prosthetic Rehabilitation of Patients Undergoing Skull Base Surgery 29. Olsen KD, Meland N, Ebersold MJ, et al. Extensive defects of the sino-orbital region: Results with microvascular reconstruction. Arch Otolaryngol Head Neck Surg. 1992;118:828–833. 30. Dexter WS, Jacob RF. Prosthetic rehabilitation after maxillectomy and temporalis flap reconstruction: A clinical report. J Prosthet Dent. 2000;83:283–286. 31. Fukuda M, Takahashi T, Nagai H. Implant-supported edentulous maxillary obturators with milled bar attachments after maxillectomy. J Oral Maxillofac Surg. 2004;62(7):799–805. 32. Triana RJ Jr, et al. Microvascular free flap reconstructive options in patients with partial and total maxillectomy defects. Arch Facial Plast Surg. 2000;2(2):91–101. 33. Taylor GI, Miller GD, Ham FJ. The free vascularized bone graft. A clinical extension of microvascular techniques. Plast Reconstr Surg. 1975;55(5):533–544. 34. Ferri J, et al. Use of the fibula free flap in maxillary reconstruction: A report of 3 cases. J Oral Maxillofac Surg. 2002;60(5):567–574. 35. Disa JJ, Cordeiro PG. Mandible reconstruction with microvascular surgery. Semin Surg Oncol. 2000;19(3):226–234. 36. Hannen EJ. Recreating the original contour in tumor deformed mandibles for plate adapting. Int J Oral Maxillofac Surg. 2006;35(2):183–185. 37. Peled M, et al. The use of free fibular flap for functional mandibular reconstruction. J Oral Maxillofac Surg. 2005;63(2):220–224. 38. Mukohyama H, et al. Rehabilitation of a bilateral maxillectomy patient with a free fibula osteocutaneous flap. J Oral Rehabil. 2005;32(7):541–544. 39. Barnouti L, Caminer D. Maxillary tumours and bilateral reconstruction of the maxilla. ANZ J Surg. 2006;76(4):267–269. 40. Funk GF, Arcuri MR, Frodel JL Jr. Functional dental rehabilitation of massive palatomaxillary defects: Cases requiring free tissue transfer and osseointegrated implants. Head Neck. 1998;20(1):38–51. 41. Peng X, et al. Maxillary reconstruction with the free fibula flap. Plast Reconstr Surg. 2005;115(6):1562–1569. 42. Disa JJ, Hidalgo DA, Cordeiro PG, et al. Evaluation of bone height in osseous free flap mandible reconstruction: An indirect measurement of bone mass. Plast Reconstr Surg. 1999;103:1371. 43. Eggbeer D, Bibb R, Evans P. Assessment of digital technologies in the design of a magnet retained auricular prosthesis. J Maxillofac Prosth Tech. 2006;9:1–4. 44. Beumer J, Roumanas E, Nishimura R. Facial defects: Alterations oat surgery to enhance the prosthetic prognosis. First Int Congress on Max Fac Prosth. 1995;18:104–107. 45. Udagama A. Urethane-lined silicone facial prosthesis. J Prosthet Dent. 1987;58:351–354. 46. Lemon JC, Chambers MS. Conventional methods of retention of facial prostheses. First Int Congress on Max Fac Prosth. 1994;1:116–119. ¨ G, Tjellstrom ¨ A, Brsnemark PI, et al. Bone anchored47. Granstom reconstruction of the irradiated head and neck cancer patient. Otolaryngol Head Neck Surg. 1993;108:334–343. 48. Hong WL, Chu SA, Dam JG, et al. Oral rehabilitation using dental implants and guided bone regeneration. Ann Acad Med Singapore. 1999;28:697–703. 49. Lee MB. Implants.
[email protected]. Support Central. April 2003. 50. Eckert SE, Desjardins RP. The impact of endosseous implants on maxillofacial prosthetics. In: Taylor TD, ed. Clinical Maxillofacial Prosthetics. Chiacgo: Quintessence, 2000:145–153. 51. Scortecci GM. Introduction to oral implantology in restorative dentistry. In: Scortecci GM, Misch CE, Benner KU, eds. Implants and Restorative Dentistry. New York: Martin Dunitz Ltd, 2001:1–25. 52. Branemark PI, Zarb G, Albrektsson T. Tissue-integrated prostheses. In: Branemark PI, ed. Osseointegration in Clinical Dentistry. Chicago: Quintessence, 1985:1–50. 53. Lemon JC, Chambers MS, Wesley PJ, et al. Rehabilitation of a midface defect with reconstructive surgery and facial prosthetics: A case report. Int J Oral Maxillofac Implants. 1996;11:101–105.
157
54. Benner KU. Morphological aspects of oral implantology. In: Scortecci GM, Misch CE, Benner KU, eds. Implants and Restorative Dentistry. New York: Martin Dunitz Ltd, 2001:26–46. 55. Schusterman MA, Reece GP, Miller MJ, et al. The osteocutaneous free fibula flap: Is the skin paddle reliable? Plast Reconstr Surg. 1992;90:787. 56. Winslow CD, Wax MK. Tissue transfer: Fibula. Available at: www.Emedicine.com. Accessed on August 28, 2003. 57. Anthony JP, Rawnsley JD, Benhaim P. Donor leg morbidity and function after fibula free flap mandible reconstruction. Plast Reconstr Surg. 1995;96(1):146–152. 58. Disa JJ, Cordeiro PG. The current role of preoperative arteriography in free fibula flaps. Plast Reconstr Surg. 1998;102(4):1083– 1088. 59. Hidalgo DA. Fibula free flap mandibular reconstruction. Clin Plast Surg. 1994;21(1):25–35. 60. Urken ML, Cheney ML, Sullivan MJ. Fibula free flaps. In: Atlas of Regional and Free Flaps for Head and Neck Reconstruction. New York, NY: Raven Press, 1995. 61. Ferguson BJ, Hudson WR, Farmer JC. Hyperbaric oxygen for laryngeal radiation necrosis. Ann Otol Rhinol Laryngol. 1987;96:1– 6. 62. Tibbles PM, Edelsberg JS. Hyperbaric-oxygen therapy. N Engl J Med. 1996;334(25):1642–1648. 63. Feldmeier JJ, Jelen I, Davolt DA, et al. Hyperbaric oxygen as a prophylaxis for radiation induced delayed enteropathy. Radiother Oncol. 1995;35:138–144. 64. Feldmeier JJ, Davolt DA, Court WS, et al. Histologic morphometry confirms a prophylactic effect for hyperbaric oxygen in the prevention of delayed radiation enteropathy. Undersea Hyper Med. 1998;25(2):93–97. 65. Feldmeier JJ, Newman R, Davolt DA, et al. Prophylactic hyperbaric oxygen for patients undergoing salvage for recurrent head and neck cancers following full course irradiation (abstract). Undersea Hyper Med. 1998;25(Suppl):10. 66. Ueda M, Kaneda T, Takahashi H. Effect of hyperbaric oxygen therapy on osseointegration of titanium implants in irradiated bone: A preliminary report. Int J Oral Maxillofac Implants. 1993;8:41–44. 67. Kornblith AB, Zlotolow IM, Gooen J, et al. Quality of life of maxillectomy patients using an obturator prosthesis. Head Neck. 1996;18:323–334. 68. Thompson RC, Mikos AG, Beahm EB, et al. Guided tissue fabrication from periosteum using preformed biodegradable polymer scaffolds. Biomaterials. 1999;20:2007–2018. 69. Miller MJ, Goldberg DP, Yasko AW, et al. Guided bone growth in sheep: A model for tissue-engineered bone flaps. Tissue Eng. 1996;2:51–59. 70. Reitemeier B, Notni G, Heinze M, et al. Optical modeling of extraoral defects. J Prosthet Dent. 2004;91(1):80–84. 71. Eppley BL. The accuracy of stereolithography in planning craniofacial bone replacement. J Craniofac Surg. 2003;14(6):934– 935. 72. Sinn DP, Cillo DE Jr, Miles BA. Stereolithography for craniofacial surgery. J Craniofac Surg. 2006;17(5):869–875. 73. Bouyssie JF, et al. Stereolithographic models derived from x-ray computed tomography. Reproduction accuracy. Surg Radiol Anat. 1997;19(3):193–199. 74. Chang PS, et al. The accuracy of stereolithography in planning craniofacial bone replacement. J Craniofac Surg. 2003;14(2):164– 170. 75. Cunningham LL Jr, Madsen MJ, Peterson G. Stereolithographic modeling technology applied to tumor resection. J Oral Maxillofac Surg. 2005;63(6):873–878. 76. Kernan BT, Wimsatt JA. Use of a stereolithography model for accurate, preoperative adaptation of a reconstruction plate. J Oral Maxillofac Surg. 2000;58(3):349–351. 77. Al-Sukhun J, et al. Stereolithography and the use of pre-adapted or fabricated plates for accurate repair of maxillofacial defects. Br J Oral Maxillofac Surg. 2006.
10 Radiobiology and Radiation Therapy of Skull Base Tumors Simon S. Lo, John H. Suh, and Eric L. Chang
break can either be achieved by a single particle track creating breaks in both strands or a combination of two individual particle tracks creating two individual single-strand breaks close to each other in space and time. A single-strand break is sublethal and repairable but double-strand breaks are not. The linear-quadratic equation below describes this model:
OVERVIEW Skull base tumors can be separated into three broad groups: (i) tumors arising from neural and vascular structures and meninges of the base of the brain (e.g., schwannoma, meningioma, pituitary adenoma, craniopharyngioma, paraganglioma); (ii) tumors arising from the skull base (e.g., chordoma, chondrosarcoma, osteosarcoma, plasmacytoma); and (iii) tumors arising from below the skull base (e.g., nasopharyngeal carcinoma, paranasal sinus tumor). This chapter covers the radiobiologic principles and radiotherapeutic aspects of skull base tumors. Other aspects of skull base tumors are being discussed in other chapters.
Surviving fraction (SF) = e −(α D+β D ) , 2
where D is the radiation dose absorbed in Gray (Gy). α is the coefficient associated with single-event cell kill and β represents the coefficient associated with cell kill as a result of interaction of sublethal events. When radiation is fractionated, the linear-quadratic equation then becomes 2 SF = Number of fractions (n) × e −(α D+β D )
RADIOBIOLOGY OF SKULL BASE TUMORS Radiation can be divided into directly or indirectly ionizing radiation. Electromagnetic radiation such as photons and gamma rays are indirectly ionizing radiation (1). Charged particles such as electrons, protons, and heavily charged ions are classified as directly ionizing radiation (1). Photons and gamma rays as well as electrons are most commonly used for radiation therapy. The availability of particle beam therapy is limited to specialized facilities. In the United States, there are currently five proton beam facilities that are actively treating patients. The biologic effect of radiation is a result of damage to DNA (1). Any kind of radiation can have direct and indirect action on the critical targets in the cells. The direct action of radiation involves direct interaction of the radiation with the critical targets in the cells leading to ionization or excitation of the atoms of the critical targets and biologic changes. This is the main mechanism of cell kill in high linear energy transfer radiation such as alpha particles and other heavily charged ions (1). Indirect action of radiation involves production of free radicals from the interaction of the radiation with the atoms or molecules in the cells. The free radicals can affect DNA and this leads to cell kill. This is the predominant mechanism of cell kill in low linear energy transfer radiation such as photons and gamma rays (1). Cell survival after a single dose of radiation is a probability function of the radiation dose (in Gray) absorbed. A typical cell survival curve consists of a shoulder at the lowdose region followed by a steeply sloped portion at high-dose regions. The shoulder is thought to represent accumulation of sublethal damages in the low-dose region and the steep portion cell kill as a result of the combination of two or more sublethal damages (Fig. 1) (1). Most cell survival curves fit into the linear quadratic model. The target for cell kill by ionizing radiation is the double helix of DNA. Therefore, double-strand breaks are required for cell kill to occur. A double-strand
Alpha–beta ratio (α:β) represents the relative contribution of the α and the β components to cell kill by radiation. α:β is different for different normal tissues and different tumor types and therefore the shapes of cell survival curves are different for different tissues. In general, tissues can be divided into two broad groups according to the values of their α:β—early- and lateresponding tissues. Normal brain parenchyma and critical structures such as the inner ear, optic pathway, pituitary, brainstem, and spinal cord are generally regarded as lateresponding tissues. Typically, early-responding tissues have much higher α:β values than late-responding tissues, which means the contribution of interaction of potentially reparable sublethal events is much smaller in early- than in lateresponding tissues. Therefore, in the setting of fractionated radiation therapy, the sparing effect on late-responding tissue by fractionation is more pronounced than that on earlyresponding tissue. The advantage of fractionated radiation therapy can be exploited if the tumor treated by radiation therapy is composed of early-responding tissue and is surrounded by late-responding tissue such as brain parenchyma. Malignant tumors typically have high α:β values (1). Benign skull base tumors such as meningiomas, acoustic schwannoma, pituitary adenomas, and craniopharyngiomas are regarded as late-responding tissues and they have low α:β values. When fractionated radiation therapy is used to treat these tumors, the sparing effects on tumor and normal brain parenchyma and other critical neural structures are similar. In circumstances where the radiation dose required to permanently control these tumors exceeds the constraints of normal brain parenchyma and critical structures, the use of radiation techniques or modalities that can achieve a very conformal radiation dose distribution and a very steep dose gradient beyond the tumor volume is indicated. 159
160
Lo et al.
and critical structures. Some of the structures at risk include brain parenchyma, brainstem, spinal cord, optic pathway, and inner ear/cochlea. In the treatment planning of fractionated radiation therapy (including IMRT, FSRT, and proton beam therapy) and SRS, the radiation dose constraints of the critical structure have to be taken into account. The commonly quoted radiation dose constraints are as follows: Brainstem—50 to 54 Gy for conventionally fractionated radiation therapy. Spinal cord—45 to 50 Gy for conventionally fractionated radiation therapy. Optic pathway—50 to 54 Gy for conventionally fractionated radiation therapy, and 8 Gy for SRS. Cochlea/inner ear—50 to 54 Gy for conventionally fractionated radiation therapy.
Figure 1
Typical cell survival curve for early- and late-responding tissues.
When a single high dose of radiation, as in stereotactic radiosurgery (SRS), is used to treat skull base tumors, the sparing effect of fractionation does not apply. From the radiobiologic standpoint, radiosurgical targets can be divided into four classes (2): Class A—target tissue is late-responding tissue and there is normal brain parenchyma within the target (e.g., arteriovenous malformation). Class B—target tissue is late-responding tissue and there is no normal brain parenchyma within the target (e.g., meningioma, acoustic neuroma, pituitary adenoma, and craniopharyngioma). Class C—target tissue is early-responding tissue and there is normal brain parenchyma within the target (e.g., low-grade glioma). Class D—target tissue is early-responding tissue and there is normal brain parenchyma within the target (e.g., highgrade glioma and brain metastasis). Class B and D targets are deemed suitable for SRS. Most skull base tumors are either class B (e.g., meningioma, acoustic neuroma, pituitary adenoma, craniopharyngioma, glomus tumor) or class D targets (e.g., chordoma, chondrosarcoma) and therefore they could be amenable to treatment with SRS. With the delivery of a single high dose of radiation that is necessary to kill the tumor, the level of cell kill will fall on the steep portion of the cell survival curve. In order to avoid collateral damage to the surrounding brain parenchyma and critical structures, the radiation dose fall-off beyond the target volume treated should be very steep such that region immediately adjacent to the target volume receives only a fraction of the prescribed dose. A stereotactic radiation delivery system can achieve this goal.
NORMAL TISSUE TOLERANCE OF CRITICAL STRUCTURES Because of the critical location of most skull base tumors, patients are at risk of developing radiation-induced complications as a result of injury to the normal brain parenchyma
Significant effort should be made to keep the radiation doses to these critical structures below the constraints to minimize the risk of radiation-induced complications. Advanced radiation techniques that can achieve very conformal isodose distribution and steep radiation dose fall-off are well suited in cases in which the tumor treated is very closed to one or more of these critical structures.
ADVANCED RADIATION THERAPY TECHNIQUES AND MODALITIES FOR SKULL BASE TUMORS There has been significant development of radiation therapy technology in the past 20 years. The skull base is a challenging area to treat with radiation therapy because two critical structures including the optic pathway and brainstem are often in close proximity. In the past, radiation therapy was delivered using two-dimensional techniques. The determination of target volume was based on bony landmarks. Because of the simplistic beam arrangement, a significant amount of brain tissue received the prescribed dose. With the advent of the three-dimensional treatment planning utilizing computerized tomography (CT) and image registration or fusion with magnetic resonance imaging (MRI), target volume coverage can be improved and more normal tissue can be spared. With further advancement of technology, it is possible to place dose constraints for various critical structures. Intensity modulated radiation therapy (IMRT) utilizes inverse planning, which allows for the placement of dose constraints on critical structures resulting in better normal tissue sparing (3). SRS is a radiation technique in which multiple external radiation beams are aimed at an imaging defined target volume to deliver a single high dose of radiation (4). A gamma R Knife unit, a linear accelerator or a cyclotron that produces protons, can be used as a platform in the delivery of SRS, which has been used to treat various skull base tumors. For skull base tumors that are larger in size and closer to critical structures, fractionated stereotactic radiotherapy (FSRT) is often used (5). The common feature of SRS and FSRT is the rapid dose fall-off beyond the target volume. Most skull base tumors are extra-axial in location and have a very sharp margin of demarcation with surrounding normal brain tissue. This makes SRS and FSRT well suited for the treatment of most skull base tumors. Because of the very favorable physical characteristics of proton beams, protons are well suited in the treatment of skull base tumors. The relatively low entrance dose and the absence of radiation dose beyond the end range known as the Bragg peak makes it possible to reduce the amount of
Chapter 10: Radiobiology and Radiation Therapy of Skull Base Tumors
radiation delivered to nontarget tissue (6). Proton beam therapy is particularly suitable for the treatment of skull base tumors that require a high radiation dose to achieve durable tumor control (6). Compared to other radiation techniques using photons, proton beam radiation therapy can result in better target volume coverage and sparing of normal structures, and a lower nontarget integral dose as a result of the Bragg peak (6).
CLINICAL APPLICATIONS OF RADIATION THERAPY (INCLUDING STEREOTACTIC RADIOSURGERY) IN SKULL BASE TUMORS Pituitary Adenomas Pituitary adenoma is a broad term that encompasses all benign primary tumors arising from pituitary glandular tissue. It can broadly be divided into nonsecretory and secretory types. Dependent upon the type of secretory cells from which they arise, secretory pituitary adenomas can be subdivided into growth hormone (GH)–secreting, prolactin-secreting, adrenocorticotrophic hormone (ACTH)– secreting, gonadotropin-secreting, and thyroid stimulating hormone–secreting tumors. The goals of treatment of pituitary adenomas are tumor control and normalization of the hypersecreted hormone (for secretory tumors). With the exception of prolactinomas, surgical resection is usually recommended because it can decompress the optic apparatus if it is compressed by the tumor and it can rapidly normalize the hypersecreted hormone in a secretory tumor (7). However, complete surgical resection is not always achievable because of involvement of critical structures (Fig. 2). Therefore, postoperative radiation therapy is often necessary to stop the growth of the residual tumor and normalize the hypersecreted hormone. Radiation therapy is also indicated for recurrent pituitary adenomas as well. There is abundant literature on the use of conventional radiation therapy for the treatment of pituitary adenoma. The typical dose used range from 45 to 54 Gy in conventional fractionation. For nonfunctioning pituitary adenoma, the
Figure 2 Coronal MRI showing extensively infiltrating pituitary adenoma centered on sella turcica and extending bilaterally into cavernous sinuses, and encasing both cavernous carotid arteries with suprasellar extension and upward bowing of optic chiasm.
161
reported 10-year progression-free survival rate after radiation therapy ranged from 80% to 98% (8–13). For secretory tumors, apart from control of tumor growth, normalization of hypersecreted hormones is one of the main goals of treatment. For GH-secreting tumors, conventional fractionated radiation therapy is efficacious in the normalization of GH. However, the GH levels tend to fall slowly after treatment and it may take years for the level to normalize. By 10 years, approximately 70% to 90% will have their GH levels normalized (9,10,14,15). In patients with lower pretreatment GH levels, normalization tends to occur earlier. Because of the metabolic effects and the cosmetic deformities associated with acromegaly, medical therapy may be given to patients before their GH levels normalize. Colleagues from St. Bartholomew’s Hospital recently reported the world’s largest series of patients with acromegaly treated with pituitary radiotherapy (16). GH level decreased to <2.5 ng/mL in 22%, 60%, and 77% of the 884 patients treated with radiotherapy for acromegaly at 2, 10, and 20 years, respectively (16). Sixtythree percent of patients had a normal insulin-like growth factor (IGF-I) level by 10 years (16). Another recent study showed similar findings (17). Table 1 summarizes the treatment outcomes of selected series. Prolactin-secreting tumors are usually managed initially by the use of bromocriptine R or cabergoline (Dostinex ). When evaluating the treatment outcomes, it is important to distinguish between pituitary stalk effect and prolactin hypersecretion. The serum prolactin level is usually lower in the case of stalk effect, which is defined as the loss of hypothalamic inhibition due to compression of the pituitary stalk. Conventional fractionated radiation therapy is usually offered to patients who cannot tolerate bromocriptine or similar medical therapy or those who develop disease progression during medical therapy. Normalization of prolactin levels occurs in 50–70% of the patients after fractionated radiation therapy (Table 1) (18–25). Patients with ACTH-secreting tumors present with Cushing disease when the tumor is still small in size, and therefore, they usually have microadenomas. Because these tumors are typically very small, they may not be detected on a regular brain MRI. Sometimes, a dynamic MRI or bilateral venous simultaneous selective venous sampling of ACTH from the inferior petrosal sinuses is done to establish a diagnosis. The metabolic effects of Cushing disease are very crippling and if longstanding, can be fatal. Therefore, the goal of treatment is rapid normalization of the ACTH, and surgical resection is usually the recommended initial treatment. Fractionated radiation therapy is recommended in patients who have residual tumor after surgery and in patients who are medically inoperable. Remission occurs in approximately 50% to 80% of the patients after fractionated radiation therapy (Table 1) (24,26–29). Nelson syndrome is described as the hypersecretion of ACTH by a pituitary adenoma in patients who have undergone bilateral adrenalectomy for Cushing disease. Unlike the usual ACTH-secreting pituitary adenomas, pituitary tumors associated with Nelson syndrome are usually macroadenomas. Surgical resection is the recommended initial treatment although surgical cure could only be achieved in less than one-third of the patients. Radiation therapy is indicated if the ACTH level remains elevated after surgery. Data on the use of fractionated radiation therapy to treat Nelson syndrome are limited. In one series, approximately 50% of patients benefited from fractionated radiation therapy in terms of clinical, biochemical, and radiographic responses (27). SRS has also been used in the treatment of pituitary adenomas. Because the optic apparatus is in proximity to the
162
Lo et al.
Table 1
Summary of Treatment Results of Fractionated Radiation Therapy for Pituitary Adenomas
Series
Type
No. of pts
Tx
Endpoint
Grigsby (10) Grigsby (9) Ludecke (15) Eastman (14)
Acromegaly Acromegaly Acromegaly Acromegaly
12 22 30 87
S + RT RT alone RT alone RT ± S
DFS DFS GH <10 ng/mL GH <5 ng/mL
Jenkins (16) Biermasz (17)
Acromegaly Acromegaly
884 36
RT ± S S + RT
GH < 2.5 ng/mL Normal IGF-I without medication Normalization of GH suppression during GTT Both of the above
Grigsby (19) Grigsby (18) Tsagarakis (21) Littley (20) Johnston (22) Rush (25) Clarke (23) Hughes (24)
Prolactinoma Prolactinoma Prolactinoma Prolactinoma Prolactinoma Prolactinoma Prolactinoma Prolactinoma
Normal PRL biochemically and clinically Normal PRL biochemically and clinically Normal PRL Normal PRL Normal PRL Normal PRL Normal PRL Normal PRL
Howlette (28) Howlette (28) Vicente (27) Littley (29) Hughes (24) Estrada (26)
Cushing ds Cushing ds Cushing ds Cushing ds Cushing ds Cushing ds
RT alone S + RT RT alone RT ± S RT ± S RT alone RT ± S S + RT RT alone RT RT for recurrent disease RT for recurrent disease RT RT RT for recurrent disease
17 28 36 58 14 10 14 19 6 21 9 14 24 40 30
Remission Remission Remission Tumor control/ remission Disease free Regression of clinical features of Cushing ds + Normal urinary cortisol excretion + Low plasma cortisol concentration in the morning after 1 mg of dexamethasone at midnight
Control (yr) 76% (10) 69% (10) 88% (5) 30%/53%/77%/89% (5/10/15/20) 22%/60%/77% (2/10/20) 75% (?) 65%/69%/71% (2.5/10/15) 40%/61%/65%/63% (3–5, 6–10, 11–15,>15) 82% (10) 93% (10) 50% (2–13) 50% (10) 43% (9) 70% (3.8) 71% (7.5) 62% (5) 50% (5) 57% (9.5) 56% (3) 61%/70% (1/2) 33% (7.7) 59% (10) 83% (3.5)
Abbreviations: No. of pts, number of patients; Tx, treatment; yr, year; S, surgery; RT, radiation therapy; DFS, disease-free survival; GH, growth hormone; IGF-I, insulin-like growth factor; GTT, glucose tolerance test; PRL, prolactin; Cushing ds, Cushing disease.
pituitary gland, a 3 to 5 mm gap is required to respect the tolerance of the optic chiasm (Fig. 3). If one anticipates that the maximum dose to the optic apparatus cannot be limited to 8 to 10 Gy while delivering an adequate dose to the tumor or target volume, fractionated radiation therapy or FSRT should be recommended instead. There are abundant data in the literature demonstrating that it is a safe and efficacious procedure for the treatment of pituitary adenomas. Again, the aims of the treatment are control of tumor growth and normalization of hormone level. SRS is very effective in the control of tumor growth in that it yields a tumor control rate of 92% to 100% (30). The tumor control rate in patients with endocrine-inactive pituitary tumors treated to lower marginal doses of 14 to 25 Gy is similar to that of patients with secretory pituitary tumors treated to marginal doses of 14 to 34 Gy. In terms of normalization of hormone level, the inconsistencies of the endpoints used in various studies render interpretation of results difficult (30). A decrease in hormone hypersecretion can occur within a few months after SRS but complete normalization can take up to 8 years. Data in the literature suggest that there is a radiation dose response in terms of normalization of hormone levels. It is also suggested that the use of antisecretory medications at the time of SRS has a negative impact on the efficacy of the procedure (30). For endocrine-inactive tumors, a dose of 15 to 20 Gy is usually used; for secretory tumors, a higher dose should be considered. For patients with acromegaly, the reported rates of endocrine cure ranged from 0% to 90% (30). However, the criteria of endocrine cure in those studies were either not defined or inconsistent. From the endocrinologists’ standpoint, an endocrine cure is usually defined as a GH level of ≤1 ng/mL
Figure 3 This Gamma Knife plan shows a dose of 25 Gy prescribed to the 50% isodose line for a growth hormone-secreting pituitary adenoma; the optic apparatus is contoured and the dose is limited to 8 Gy.
Chapter 10: Radiobiology and Radiation Therapy of Skull Base Tumors
and a normal IGF-I. A comprehensive review was done examining the outcomes of SRS for the treatment of acromegaly (30). Among the studies using the criteria of endocrine cure closest to the criteria used by the endocrinologists, the rate of endocrine cure ranged from 20% to 82% (30). The author commented that this wide variation might be a result of the different percentage of patients receiving antisecretory medications during SRS. For prolactinomas, SRS results in an endocrine cure rate approximately 30% and a significant reduction of PRL levels in 29% to 100% of the patients (30). The variation in results may be related to different proportions of patients receiving antisecretory medications. The post-SRS stalk effect may cause the PRL to be slightly elevated even when the hypersecretion is well controlled (30). For ACTH-secreting tumors, the reported endocrine cure rates after SRS ranged from 10% to 100% (30). However, some of the studies did not specify the criteria of endocrine cure and other studies used various different criteria. Among the studies with defined criteria of endocrine cure, the rates of endocrine cure ranged from 28% to 100% (30). Data on SRS for the treatment of Nelson syndrome is scarce in the literature. The endocrine cure rates ranged from 0% to 36% (30). In the recent years, data on the use of FSRT for the treatment of pituitary adenomas have been emerging. The prescribed dose ranged from 45 to 52.2 Gy in conventional fractionation. The tumor control rates ranged from 85% to 98% (7). The reported complication rates were low. Longer follow-up is required to determine the long-term efficacy and toxicity of FSRT for the treatment of pituitary adenomas.
Acoustic Schwannoma Acoustic schwannomas are slow growing benign tumors and are usually managed with microsurgery, SRS, FSRT, or observation. Since acoustic schwannoma are slow growing tumors, some physicians favor close observation of smaller tumors. However, tumor progression can result in permanent loss of hearing function and can increase the risk of complications associated with surgery or SRS/FSRT. Comparisons have been made between microsurgery and SRS; it is demonstrated that SRS yields similar tumor control rates but a lower incidence of cranial nerve deficits although very long-term data are lacking for SRS (31–34). Acoustic schwannomas are regarded as Class B targets, which are ideal targets for SRS (Fig. 4). There are abundant data in the literature on the utilization of SRS for the treatment of acoustic neuroma. There has been an evolution of the treatment techniques, imaging, and radiation dosing for SRS in the management of acoustic schwannoma. In the past, more rudimentary planning was used and plans were typically less conformal given earlier computer algorithms and CT-based
Figure 4 Axial MRI showing acoustic schwannoma involving the right internal auditory canal and extending into the right cerebellopontine angle with central necrosis following stereotactic radiation therapy.
planning. Modern treatment planning entails the use of larger number of shots to improve the conformality around the tumor (35). The availability of inverse planning in linear accelerator-based SRS has also improved the conformality index. In regard to dosing, the prescribed dose for acoustic schwannoma has been lowered over the past one to two decades. In the earlier studies, marginal doses in the range of 16 to 20 Gy were used (35–37). Although the reported tumor control rates were excellent, the incidence of radiationinduced hearing loss and cranial nerve deficits (mainly trigeminal and facial nerves) was substantial prompting physicians to lower the prescribed dose for SRS (35–38). More recent series utilizing a lower prescribed dose of 12 to 13 Gy did not show inferior tumor control rates (37–48). In terms of hearing preservation, a much higher proportion of patients retained serviceable hearing. There is also a lower incidence of trigeminal and facial nerve injury from SRS. Table 2 summarizes the treatment results of selected SRS series. One of the largest series of 829 acoustic schwannoma patients treated with Gamma Knife-based SRS from University of Pittsburgh reported a 10-year tumor control rate of 97% (37). Other series reported similar high tumor control rates (38–42,44,47,48). Another study from the same institution examined the treatment outcomes of patients with acoustic schwannoma treated with a reduced dose of 12 to
Table 2 Summary of Results of Selected SRS Series for Acoustic Neuroma Series Flickinger (38) Inoue (39) Petit (40) Paek (41) Iwai (44) Weber (45) Combs (49)
163
No. of pts
Modality/dose
Tumor control (Follow-up)
Hearing preservation
V/VII nerve toxicity
313 18 47 25 51 88 26
GK/12–13 Gy GK/10–12 Gy GK/12 Gy GK/12 Gy GK/12 Gy Protons/12 Gy LINAC/13 Gy
98.6% (6 yr) 93.3% (6–13 yr) 96% (3.6 yr) 92% (45 mo) 92% (60 mo) 93.6% at 5 yr (38.7 mo) 91% at 10 yr (110 mo)
78.6% 80% 88% 52% 59% 33% 55% at 9 yr
4.4%/0 0/0 0/4% 5%/0 0/0 10.6%/8.9% 8%/5%
Abbreviations: SRS, stereotactic radiosurgery; No. of pts, number of patients; V/VII nerve toxicity, trigeminal/facial nerve toxicity; GK, Gamma Knife; LINAC, linear accelerator.
164
Lo et al.
13 Gy and demonstrated that tumor control was not compromised (38). With a median follow-up of 24 months, the rates of preservation of facial nerve function, trigeminal nerve function, and serviceable hearing were 100%, 95.6%, and 78.6%, respectively (38). A series from University of Maryland utilizing a reduced dose of 12 Gy for SRS for acoustic schwannoma showed a tumor control rate of 96% with a median follow-up of 3.6 years (40). Eighty-eight percent of the patients retained hearing. No patient experienced facial numbness. Transient facial numbness was reported in 4% of the patients (40). Another series from Japan utilizing a reduced dose of 12 Gy showed a similarly high tumor control rate with none of the 51 patients developing any facial numbness or weakness (44). Colleagues from Harvard University demonstrated that proton beam SRS yielded a 5-year tumor control rate of 95.3% with a median follow-up of 38.7 months (45). The prescribed dose was 12 cobalt Gray equivalent (CGE). The trigeminal and facial nerve toxicity rates were 10.6% and 8.9%, respectively (45). The hearing preservation rate was 33.3% (45). The group from University of Heidelberg reported the long-term treatment outcomes of patients with acoustic schwannoma treated with linear accelerator-based SRS. A median dose of 13 Gy was prescribed to the 80% isodose line. With a median follow-up of 110 months, the 5- and 10-year tumor control rates were 91% (49). The rates of radiation-induced trigeminal neuralgia and facial weakness were 8% and 5%, respectively (49). The hearing preservation at 9 years was 55% (49). Other series showed similar results for tumor control and trigeminal and facial nerve toxicities and varying hearing preservation rates. The follow-up times of most of these studies are relatively short and additional follow-up is required to determine the long-term efficacy and toxicity of SRS using a reduced dose approach. FSRT using conventional or hypofractionation has also been used to treat acoustic neuromas (43,50–60). In particular, patients with large acoustic schwannoma not suitable for SRS may be offered FSRT. The largest body of experience with conventional FSRT came from University of Heidelberg, Germany. A total of 106 patients with acoustic schwannoma were treated with FSRT (50). The prescribed dose was 57.6 Gy in conventional fractionation. With a median follow-up of 48.5 months, the 5-year actuarial tumor control was 93% (50). The toxicities associated with the treatment were low with 3.4%, 2.3%, and 6% developing trigeminal nerve, facial nerve, and vestibulocochlear nerve complications, respectively (50). Investigators from University of California, Los Angeles demonstrated similar results with the use of FSRT for the treatment of acoustic schwannoma (57). Investigators from Loma Linda University also reported their treatment results of fractionated proton beam radiotherapy for the treatment of acoustic neuroma. The prescribed dose was 54 CGE or 60 CGE (for patients with no serviceable hearing) in 30 fractions (61). With a median follow-up of 34 months, the tumor control rate was 100% (61). No trigeminal or facial nerve toxicity was reported. The hearing preservation rate was 31% (61). The group from Thomas Jefferson University also showed similar tumor control rates and incidence of cranial nerve toxicity (trigeminal and facial nerves) for patients with acoustic schwannoma treated with FSRT in conventional fractionation (43). In patients with sporadic tumors, the probability of retaining serviceable hearing was 81%. Hypofractionated regimens have also been described for FSRT (51–54,62,63). The regimens used include 3 Gy × 10, 5 Gy × 4, 5 Gy × 5, 4 Gy × 5, 6 Gy × 3, and 7 Gy × 3. Regardless of the regimen used, the tumor control rate ranged from 94% to 100%. The rates of trigeminal and facial nerve toxicity were low. Approximately
one-fourth to one-third of the patients developed complications related to hearing. The median follow-up intervals for the FSRT studies ranged from 21 to 48.5 months. It is still too early, due to recent SRS dose reductions, to draw conclusions as to whether FSRT or SRS with lowered prescribed doses has the advantage in terms of the therapeutic ratio. However, in patients not eligible for SRS based on tumor size, FSRT represents a reasonable treatment alternative.
Meningioma Skull base meningiomas can be addressed with skull base surgical techniques. However, the critical structures that reside in the region of the skull base can prevent complete resection from being performed safely, leading to increased risk of tumor progression. To improve tumor control, postoperative radiation therapy is indicated for patients who have subtotally or minimally resected newly diagnosed or recurrent skull base meningiomas (64). There are abundant data in the literature supporting the use of conventional fractionated radiation therapy in the management of patients with subtotally or minimally resected newly diagnosed or recurrent skull base meningiomas (64). The majority of the series included meningiomas in skull base as well as non–skull base locations. The typical prescribed dose ranges from 50 to 54 Gy for benign meningiomas in most cases. For patients who receive radiation therapy after subtotal resection, it is reasonable to expect a 10-year local control rate of at least 70% to 80% based on the long-term results reported in the literature. For patients with unresectable disease, radiation therapy provides a degree of tumor control and symptomatic relief. In the series from Royal Marsden Hospital, United Kingdom, the reported 5-, 10-, and 15-year disease-free survival rates for patients with unresectable meningiomas were 53%, 47%, and 47%, respectively (65). For patients with recurrent meningiomas, there is some suggestion that the outcomes are not compromised if radiation therapy is given at the time of recurrence compared to the outcomes of patients who receive immediate radiation therapy after subtotal tumor resection (66). Because of the presence of critical structures—such as the optic apparatus, the cranial nerves in the cavernous sinus and the brainstem—progression of a skull base meningioma can result in significant neurological morbidity. This question is best answered in a clinical trial setting. The European Organisation for Research and Treatment of Cancer (EORTC) is conducting a trial randomizing patients with subtotally resected or biopsied World Health Organization grade 1 cerebral meningiomas to observation or postoperative external beam radiation therapy or SRS (EORTC 26,021-22,021). Because meningiomas have a sharp margin of demarcation with normal brain parenchyma, they are excellent targets for highly precise radiation techniques such as IMRT, SRS, FSRT, and proton beam therapy [Fig. 5(A)]. The main advantage of using these techniques is the capability of minimizing radiation dose delivery to the areas outside the target volume. IMRT can produce highly conformal isodose distribution around a target volume while minimizing radiation dose delivery to surrounding critical structures or organs [Fig. 5(B)]. It is well suited for treatment of targets with complex shapes such as meningiomas. Colleagues from Baylor College of Medicine treated 40 meningioma patients (32 with skull base tumors) with IMRT to a dose of 50.4 Gy. With a median follow-up of 30 months, the 5-year tumor control was 93% (67). Acute and late complications occurred in 2.5% and 5% of the patients, respectively (67). Another study from University of Heidelberg examined the
Chapter 10: Radiobiology and Radiation Therapy of Skull Base Tumors
(A)
165
(B)
Figure 5 (A) Axial MRI T1 postcontrast image showing hyperostosing en plaque meningioma of the right greater sphenoid wing with extension into the lateral portion of the right orbital apex and right parasellar region. (B) Intensity modulated radiation therapy plan to deliver 50.4 Gy in 30 fractions to this meningioma.
treatment outcomes of 20 patients with skull base meningiomas treated with IMRT. The prescribed dose was 57.6 Gy. With a median follow-up of 36 months, the tumor control rate was 100% with 25% of the tumors treated showing tumor shrinkage (68). The incidence of late complications was 10% (68). A recent study of 35 patients with 37 meningiomas treated with IMRT at Cleveland Clinic Foundation showed similar findings. The prescribed dose was 50.4 Gy. With a median follow-up of 19.1 months, the 3-year local control was 97% and no late complications were observed (69). Overall, all these studies had relatively short follow-up intervals ranging from 19.1 to 36 months. Extended follow-up is required to determine whether IMRT can improve the therapeutic ratio compared to conventional fractionated radiation therapy. Data on SRS for the treatment of skull base meningioma have emerged in the literature over the last two decades (64). Meningiomas are category 2 targets with a sharp margin of demarcation from normal brain parenchyma, and therefore, they are ideal targets for SRS (2). The typical prescribed dose was 12 to 18 Gy. Numerous reports have demonstrated the high efficacy and the low toxicity associated with SRS for the treatment of skull base meningioma (70–74). Colleagues from Mayo Clinic reported the results of 49 patients with cavernous sinus meningioma treated with Gamma Knife-based SRS. The mean margin dose was 15.9 Gy. With a median follow-up of 58 months, the tumor control rate was 100% (70). Twenty-six percent of patients with preexisting diplopia or facial numbness/weakness had improvement of their symptoms. Ten percent of patients developed worsening of trigeminal nerve function after treatment (70). One hundred fifty-nine patients with cavernous sinus meningioma were treated at University of Pittsburgh with Gamma Knife-based SRS. The median margin dose was 13 Gy. Neurologic status remained stable or improved after treatment in 91% of patients (72). Tumor progression occurred in 6% of patients. Adverse effect occurred in 6.7% of patients. The 5- and 10-year tumor control rates were both 93.1%. At the same institution, 62 patients with petroclival meningiomas were treated with Gamma Knifebased SRS (75). The median margin dose ranged from 11 to 20 Gy. With a median follow-up of 37 months, tumor progression occurred in 8% of the patients (75). New cranial nerve
deficits occurred in 8% of patients. Other SRS series, either Gamma Knife- or linear accelerator-based, showed similar tumor control and toxicity rates for skull base meningioma (71,73,74). There does not appear to be any significant difference between Gamma Knife-based and linear acceleratorbased SRS in terms of treatment outcomes. FSRT can be offered to patients with skull base meningiomas not suitable for treatment with SRS either due to the size limit or the proximity to critical structures like the optic apparatus. It is also a reasonable treatment alternative to either fractionated radiation therapy or SRS. Data on FSRT for the treatment of skull base meningiomas have emerged over the last 10 years. Colleagues from Royal Marsden Hospital, United Kingdom reported their preliminary treatment outcomes of patients with mostly skull base meningiomas treated with FSRT in two separate publications (76,77). The prescribed dose was 50 to 55 Gy in conventional fractionation. None of these patients developed any tumor recurrence. The group from University of Heidelberg reported the results of one of the largest FSRT series for skull base meningiomas. The prescribed dose was 56.8 Gy in conventional fractionation. The tumor control rate was 98.3% and the complication rate was 1.6% with a median follow-up of 35 months (78). Other studies using the same approach showed similar tumor control and complication rates (79–81). However, the follow-up times of those studies were relatively short. Given the indolent nature of benign meningiomas, a prolonged follow-up of these patients is necessary to determine the long-term outcomes. Secondary to the lack of wide availability of proton beam facilities worldwide, protons have not been routinely used in patients with skull base meningiomas. Proton beam therapy has been used as the sole modality or combined with photon beam therapy for the treatment of skull base meningioma. Investigators from Institut Curie, France treated 51 patients with skull base meningioma with a combination of photon and proton beam therapy. With a median follow-up of 25.4 months, the 4-year local control was 98% (82). The prescribed dose was 60.6 CGE. Investigators from South Africa treated 23 patients with skull base meningiomas with either hypofractionated stereotactic proton beam radiation
166
Lo et al.
therapy (31.5 CGE in 3 fractions) or stereotactic proton beam radiation therapy (54 CGE in 27 fractions to 61.6 CGE in 16 fractions). With a median clinical and imaging follow-up of 40 and 31 months, respectively, 88% of the patients treated with hypofractionated and 100% of the patients treated with conventionally fractionated stereotactic proton beam radiation therapy achieved tumor control documented by imaging (83). Other proton beam radiotherapy series for meningiomas that included tumors of all locations showed similar outcomes (84–86).
Craniopharyngioma Craniopharyngiomas are benign tumors arising from the Rathke’s pouch. Although they are slow growing and well circumscribed, their frequent involvement of adjacent structures such as the optic apparatus, pituitary stalk, hypothalamus, and major blood vessels can contribute to significant morbidity and can render safe complete surgical resection difficult. Surgical intervention is the standard initial therapy for the purpose of tissue diagnosis and decompression. Attempts at achieving complete surgical resection can be associated with significant morbidity. Therefore, limited surgical resection and postoperative radiation therapy can be used to achieve satisfactory rates of tumor control. Conventional fractionated radiation therapy has been shown to be effective in the setting of postoperative treatment as well as for salvage treatment. Because craniopharyngiomas are very well circumscribed tumors, they are very suitable targets for advanced radiation therapy techniques such as IMRT, SRS, FSRT, and proton beam therapy. In patients with a cystic lesion, intralesional phosphorus-32 (P-32) may be used to treat the tumor. Conventional radiation therapy may be offered in two different settings—for initial treatment of a subtotally resected tumor and for salvage treatment of recurrence after surgical resection. Data in the literature showed that radiation therapy is effective in the reduction of the risk of recurrence in both settings (87). The most commonly prescribed dose is 50 to 54 Gy. For patients who receive immediate postoperative radiation therapy, the 10- and 20-year local control or progression-free survival ranged from 57% to 89.1% and from 54% to 79%, respectively (87). For patients who receive postoperative radiation therapy as salvage treatment, similar outcomes were observed (87). This raises the question as to whether delayed instead of immediate radiation therapy should be employed, especially in young children. The inverse planning nature of IMRT enables radiation oncologist to select dose constraints to critical structures and is well suited for the treatment of craniopharyngiomas that are well circumscribed. However, because it is a recently developed treatment technique, long-term data on its efficacy and toxicity for the treatment of craniopharyngioma are lacking. Recently, there is renewed interest in the use of proton beam therapy for the treatment of brain and skull base tumors, especially in children. The physical characteristics of protons can yield an improved isodose distribution over what can be achieved with photons (Fig. 6). This is especially important in young children who receive radiation therapy for brain or skull base tumors. Colleagues from Harvard University reported their experience of combining photons and protons for the treatment of craniopharyngioma. A total of 15 patients, 5 children (median age, 15.9 years) and 10 adults (median age, 36.2 years), were treated with 160 MeV proton beam therapy either for the entire course of treatment or combining with photon beam therapy. The median dose given was 56.9 CGE. With a me-
Figure 6 A proton beam radiotherapy plan showing a dose of 50.4 CGE in 28 fractions prescribed to this large craniopharyngioma.
dian follow-up of 11 years for the 11 surviving patients, the 5- and 10-year local control rates were 93% and 85%, respectively (88). The 10-year overall survival rate was 72%. None of the 10 adults treated had any change of functional status or working ability. One of the five children had learning difficulties comparable to pre-radiation therapy level; the remaining had professional achievements. SRS (Gamma Knife-based or linear accelerator-based) has been utilized to treat craniopharyngiomas. Because of the proximity of most craniopharyngiomas to the optic apparatus, there would be a risk of radiation-induced visual disturbance if the radiation dose exceeds the tolerance level. Typically, a 3 to 5 mm gap between the tumor and optic apparatus is needed to limit the radiation dose to the structure. Most SRS series showed a local control rate of 86% to 90% (87). In one Swedish series, the rates of tumor progression were 85% and 33% for tumors receiving <6 Gy and ≥6 Gy, respectively (89). The series showing good local control rates all had relatively short follow-up times of less 3.5 years (87). Longer follow-up is needed to determine the efficacy and toxicity of SRS for craniopharyngiomas. FSRT has also been used to treat children with craniopharyngiomas. Preliminary results from various studies showed that it was safe and efficacious (90–93). In one of the large studies from University of Heidelberg, 26 patients were treated with FSRT with a median dose of 52.2 Gy (range 50–57.6 Gy). With a median follow-up of 43 months, the 10year tumor control probability and survival rate were 100% and 83%, respectively (90). Out of the 18 patients at risk, three developed endocrine deficits (90). For patients with cystic tumors, intracavitary irradiation can be used as the primary treatment or adjuvant to surgery or external beam radiation therapy. In one of the large series from University of Pittsburgh, 49 patients were treated with colloidal chromic P-32 using a stereotactic approach. The prescribed dose to the cyst wall was a dose of
Chapter 10: Radiobiology and Radiation Therapy of Skull Base Tumors
167
200 to 250 Gy. The actuarial 5- and 10-year local control rates were 76% and 70%, respectively (94). Forty-eight percent of the patients treated had improvement or stabilization of their vision. Three of the 49 patients (6%) developed permanent radiation injury related to P-32 therapy (94). In another series of 62 patients with craniopharyngioma treated with yttrium-90 (Y-90) or P-32, the actuarial 5- and 10-year survival rates were 80% and 64%, respectively (95). After treatment, three patients developed amaurosis, one developed increased visual field cut, and another one developed oculomotor nerve deficit. Three patients developed hypothalamic-pituitary dysfunction after therapy (95).
Glomus Tumor Glomus tumors, also known as nonchromaffin paragangliomas, arise from neuroepithelial cells. The most common locations are carotid body and the temporal bone. Tumors arising along the superior bulb of the internal jugular vein and along the tympanic branch of glossopharyngeal nerve are called glomus jugulare and glomus tympanicum, respectively. They are low-grade malignant tumors where the main pattern of spread is local, and nodal and distant metastases are rare. Because these tumors are hypervascular, they are usually demonstrated very well on CT with contrast. They also have a characteristic appearance on MRI. Angiography is used to evaluate patients preoperatively. The standard treatment is surgical resection or radiation therapy. However, in cases where the risk is regarded as too high for surgical resection, radiation therapy is often offered (Fig. 7). Radiation therapy should also be considered after subtotal resection. The goals of the treatment include tumor and symptomatic control. Data in the literature showed that fractionated radiation therapy is an effective treatment for glomus tumors (96–103). In terms of target delineation, CT, MRI, and angiography can be used. Since the risk of nodal metastasis is low, the clinical target volume does not typically include the nodal regions. Usually, a prescribed dose of 45 to 50 Gy in conventional fractionation is used. Springate and colleagues did a comprehensive review of all the articles published on radiation therapy for glomus tumors from 1965 to 1988. The tumor control was 93% and the risk of complication was 2% to 3% (104). Table 3 summarizes the treatment outcomes of selected series for fractionated radiation therapy for glomus tumors. SRS has also been used to treat glomus tumors. SRS can yield a very conformal isodose distribution and a very rapid dose fall-off beyond the target volume. Multiple series in the literature showed tumor control rates of 63% to 100% and symptomatic control rates of 25% to 60% (105–111). Complications occurred in 4% to 40% of the patients. One note of caution is that these series only had follow-up intervals of 20.5
Figure 7 Coronal MRI T1 postcontrast MRI showing massive skull base/jugular foramen glomus tumor centered on the right occipital bone and clivus to include regions of the hypoglossal canal.
to 51 months. Because late recurrence can occur in glomus tumors, a much longer follow-up is required to determine the efficacy of SRS in the management of glomus tumors.
Chordoma Chordomas are rare tumors that originate from the rests of the embryonal notochord. Approximately one-third of these tumors occur in the skull base. They are locally aggressive tumors that are most often treated with maximal safe resection and postoperative radiation therapy. Although gross total resection is the goal of surgery, there are instances where this is not achievable (Fig. 8). Because of the aggressive nature of the tumor, there is a high risk of local recurrence after surgical resection. Postoperative radiation therapy is employed to reduce the risk of local recurrence. The most common cause of death is uncontrolled local tumor progression, so local tumor control is of utmost importance (6). Unfortunately, the location of these tumors renders the delivery of an adequately high dose of radiation to the area difficult even with highly conformal techniques with photon beam therapy, subsequently resulting in nondurable tumor control. The treatment outcomes with the use of conventional fractionated radiation therapy have been disappointing. The typical prescribed dose for conventional fractionated radiation therapy is 50 to 60 Gy (6). This strategy resulted in a recurrence rate of 50% to 100% (6). Other strategies to intensify the radiation
Table 3 Summary of Treatment Outcomes of Patients Treated with Conventional Fractionated Radiation Therapy for Glomus Tumors Series Pemberton (102) Krych (103) Wang (97) Konefal (99) Sharma (100) Hinerman (98) Cummings (101) Powell (96)
No. of pts
RT dose
Follow-up
Tumor control
49 23 15 22 40 43 45 46
37.5–50 Gy in 15–16 days 45 Gy 29–67.5 Gy 46–55 Gy 40–50 Gy 37.7–60 Gy 35 Gy in 3 wk 35–66 Gy
7.4 yr 13.4 yr 5–33 yr 10.5 yr 13 yr 11.1 yr 10 yr 9 yr
92% 100% 80% 91% 83% 93% 93% 85%
Abbreviations: No. of pts, number of patients; RT dose, radiotherapy dose.
168
Lo et al.
Figure 8 Axial MRI postcontrast showing chordoma centered at the sella region with extension into the left side of the midbrain and pons, and involvement of both cavernous sinuses, and encasement of the left internal carotid artery.
dose to skull base chordomas such as SRS and FSRT have also been used. In the SRS series from Mayo Clinic, with a median follow-up of 4.8 years, the 2- and 5-year tumor control rates were 89% and 32%, respectively (112). The median margin dose was 15 Gy. Patients who had prior fractionated radiotherapy had a higher risk of radiation-induced complications. Other studies also suggested that SRS may be an effective treatment for skull base chordomas (113,114). FSRT has also been used to treat skull base chordomas. The group from University of Heidelberg treated patients with skull base chordomas with FSRT. The prescribed dose was 66.6 Gy. The 2- and 5year local control rates were 82% and 50%, respectively (115). Secondary to the absence of exit dose from protons, a very conformal distribution around the target volume can be achieved in the three-dimensional space (6). This enables a higher dose of radiation to be delivered to the tumor (chordoma), thereby improving the chance of tumor control. A large body of experience has been accumulated over the years on the use of proton beam therapy for the treatment of skull base chordoma. In the United States, the proton beam experience for skull base chordomas came mainly from Lawrence Berkeley National Laboratory, Loma Linda University, and Harvard University. Investigators from Harvard University reported one of the largest experiences in using proton beam therapy (5 fractions per week, 4 with protons and 1 with photons) for the treatment of skull base chordoma and low-grade chondrosarcoma. The prescribed dose was 66 to 83 CGE in conventional fractionation. A relative biologic effectiveness of 1.1 was used. Out of the 519 patients, 290 had chordoma. With a median follow-up of 41 months, the 5- and 10-year local recurrence-free survival was 64% and 42%, respectively (6). Severe complications occurred in 8% of the patients. Colleagues from Loma Linda University treated 58 patients with skull base chordoma (n = 33) and low-grade chondrosarcoma (n = 25) with proton beam therapy. The median dose was 64.8 to 79.2 CGE (6). With a mean follow-up period of 33 months, eight (24%) of 33 patients with chordomas experienced recurrence. The 3-year local control rate was 67%. The 5-year overall survival rate was 79%. A group from Institut Curie treated 100
patients with base of skull or cervical spine chordomas with fractionated radiotherapy combining photons and protons. A prescribed dose of 67 CGE was given. With a median follow-up of 31 months, a total of 25 patients developed recurrence. The 2- and 4-year tumor control rates were 86.3% and 53.8%, respectively (116). Other proton beam radiotherapy series have shown similar results (117–122). Given the more favorable outcomes associated with the use of proton beam radiotherapy compared to those achieved with photon beam irradiation alone, proton beam radiotherapy should be considered in most patients with skull base chordomas. Carbon ion therapy combines the physical advantages of protons with the differential increase in relative biologic effectiveness (123). It is currently not available in the United States and is only available in even more limited facilities worldwide. The investigators from University of Heidelberg reported that a 3-year tumor control rate of 81% was achieved for patients with chordomas and no excessive toxicity was observed (123). At German Ion Research Center (GSI), a phase I/II trial of carbon ion therapy for the treatment of skull base chordomas and chondrosarcomas has been completed in 2001. With a median follow-up of 32 months, the 4-year actuarial local control for chordomas was 74% (124). The prescribed dose was 60 GyE at 3 GyE daily, 7 days a week. These preliminary results appeared to be as favorable as those achieved with proton beam therapy. However, a longer follow-up is required to determine the long-term efficacy and toxicity associated with the treatment.
Chondrosarcoma Chondrosarcomas constitute 10% to 20% of all malignant bone tumors. Those occurring in the head and neck region most commonly arise from the sphenopetrosal and the spheno-occcipital synchondroses, the nasal cavity, and the paranasal sinuses. Complete surgical resection is the mainstay of treatment. Unfortunately, the location of the tumor often precludes the possibility of safe maximal resection. Postoperative radiation therapy is required to reduce the risk of recurrence. A high dose of radiation is required to achieve permanent local control. Similar to skull base chordomas, skull base chondrosarcomas are subject to the radiation dose constraints of surrounding critical or normal structures and the delivery of a high dose of conventional photon beam radiotherapy is often not possible. Advanced radiation techniques such as SRS and FSRT have been used to treat patients with skull base chondrosarcoma. In the series from Mayo Clinic, all the four patients with skull base chondrosarcoma treated with Gamma Knife-based SRS achieved tumor control with a median follow-up of 4.8 years (112). Similarly, good results were shown in the study from University of Pittsburgh (113,114). At the University of Heidelberg, eight patients were treated with FSRT for skull base chondrosarcoma. The prescribed dose was 64.9 Gy. None of the patients treated developed recurrence after 5 years (115). There were several proton beam radiotherapy studies pertaining to management of skull base chondrosarcoma. The largest experience came from Harvard University where 200 patients were treated. The prescribed dose was 72.1 CGE. The 5-year local control was 99% (125). Other proton beam studies utilizing similar radiation dose levels showed similar tumor control rates (117–120). Carbon ion therapy has been used to treat patients with skull base chondrosarcomas. Fifty-four patients were treated for skull base low- or intermediategrade chondrosarcoma at GSI in Germany. A dose of 60 CGE (weekly dose 7 × 3 CGE) was prescribed to the tumor volume (126). With a median follow-up of 33 months, the tumor
Chapter 10: Radiobiology and Radiation Therapy of Skull Base Tumors
control rates at 3- and 4-years were 96.2% and 89.8%, respectively (126). The grade 3 late-toxicity rate was 1.9% (126). Carbon ion therapy appears to yield similar tumor control and low toxicity rate compared to proton beam radiotherapy on short-term follow-up.
Nasopharyngeal Carcinoma Nasopharyngeal carcinoma (NPC) is relatively rare in the western hemisphere. Because of its tendency to involve the fissures and the foramina in the skull base, all of these areas are included in the radiation field. NPC also carries a high risk of bilateral neck node metastases, the whole neck has to be included in the radiation therapy field even if there is no clinical evidence of cervical node involvement (127). Unlike other head and neck cancer, NPC is managed primarily with radiation therapy with or without chemotherapy and the role of surgery is limited to neck dissection if there is residual disease after radiation therapy and to salvage nasopharyngectomy for recurrent disease after primary radiation therapy. The typical radiation dose for gross disease is 70 Gy although a dose of 66 Gy have been used in patients with nonkeratinizing carcinoma and undifferentiated carcinoma with good results. Multiple phase III trials have established the role of concurrent chemoradiation therapy in the management of NPC (128,129). A recent meta-analysis showed that chemotherapy did lead to a small but significant benefit for overall survival and event-free survival in locally advanced NPC and the advantage was seen mainly when chemotherapy was given concurrently with radiotherapy (128). Because of the complex shape of the target volume to be treated and the close proximity of the portion of the target volume in the skull base to various critical structures, the use of advanced technology such as IMRT can help optimize the therapeutic outcome (130,131). Data on the use of IMRT for the treatment of NPC are emerging over the past 5 years. Preliminary results reported in various publications showed good tumor control rates and capability of normal tissue sparing. In the study from the University of California, San Francisco, sixty-seven patients with NPC were treated with IMRT. The 4-year local progression-free, local-regional progression-free, and distant metastasis-free rates were 97%, 98%, and 66%, respectively (132). The 4-year overall survival was 88% (132). The toxicity rate was low. Other studies also showed favorable results (133,134). The ongoing Radiation Therapy Oncology Group 0615 trial involves concurrent chemoradiotherapy using 3-D conformal radiotherapy or IMRT plus bevacizumab for locally or regionally advanced nasopharynx cancer.
Paranasal Sinus Tumors Paranasal sinus tumors are rare malignancies. Histologic subtypes include epithelial tumors such as squamous cell carcinoma, undifferentiated carcinoma, malignant salivary gland tumors (adenocarcinoma, adenoid cystic carcinoma, and mucoepidermoid carcinoma), and esthesioneuroblastoma, melanoma, rhabdomyosarcoma, plasmacytoma, and lymphoma. For lymphoma and plasmacytoma, low-tomoderate dose radiation therapy (30–50 Gy) with or without chemotherapy (for lymphoma, depending on histology) is the mainstay of treatment. For other histologies, surgical resection should be considered. Radiation therapy with or without chemotherapy is often combined with surgical resection for the management of paranasal sinus tumors to enhance tumor
169
control. Typically, a higher dose of radiation is required. This presents a therapeutic challenge because the target volume to be treated is often in close proximity to critical structures such as the eyes and the optic apparatus. For patients with squamous cell carcinoma of the paranasal sinus, surgical resection is typically followed by postoperative radiation therapy (135,136). However, patients often present with very advanced disease that renders surgical resection not feasible. Primary radiation therapy is offered in that setting. Data in the literature showed that the treatment results were suboptimal. The reported 5-year tumor control rate ranged from 41% to 78% (136). The 5-year overall survival rate ranged from 27% to 55% (136). High nodal failure rates have been reported when elective neck irradiation is not given. Sinonasal undifferentiated carcinoma is a rare malignancy that runs a very aggressive course. If feasible, patients are frequently managed with surgical resection with or without adjuvant or neoadjuvant radiation therapy, chemotherapy, or both. If the disease is too advanced for surgical extirpation, primary radiation therapy with or without chemotherapy can be offered (137). The reported survival rates after treatment ranged from 20% to 63% (137). Given the high risk of cervical nodal metastasis, elective neck irradiation is usually recommended (137). Esthesioneuroblastoma is usually managed with surgical resection followed by postoperative radiation therapy. For patients who are not candidates for surgery, primary radiation therapy is given (138). Overall, the treatment outcomes associated with primary radiation therapy alone are poor and a complete resection followed by postoperative radiation therapy is associated with much better outcomes (139–141). Late recurrences have been observed in a significant proportion of patients after treatment, so the importance of long-term follow-up of these patients cannot be overemphasized (142,143). Because a high dose of radiation (60–70 Gy) is required for the treatment of paranasal sinus epithelial tumors, there is a substantial risk of retinopathy and optic neuropathy. To decrease the risk of collateral damage, different strategies have been used. The use of 1.2 Gy twice daily regimen have been used by the group at University of Florida (137). With the availability of advanced technology, a highly conformal radiation dose distribution around the target volume can be readily achieved. The use of intensity-modulated radiation therapy for the treatment of paranasal sinus tumors have been reported in the literature and the preliminary results appeared to be promising in terms of reduction of toxicity (144–146).
RADIATION TOXICITY The complications related to radiation therapy for skull base tumors are dependent upon the critical structures treated and the radiation dose delivered. The skull base is a challenging location for radiation delivery because various critical structures are located in the area. Structures at risk include normal brain parenchyma, optic apparatus, cranial nerves, major arteries, inner ear, brainstem, spinal cord, pituitary, and hypothalamus. For very young patients or children receiving fractionated radiation therapy for the treatment of skull base tumors, they are especially at risk of developing neurocognitive deficits and growth impairment. As most skull base tumors are benign tumors that run a long clinical course, there is a concern for radiation-induced tumors.
170
Lo et al.
SUMMARY Skull base tumors present a challenge to neurosurgeons, otolaryngologists, and radiation oncologists. The best management approach for patients with skull base tumor is interdisciplinary treatment planning. Before any treatment is initiated, it is important for the treatment team to plan ahead of time the anticipated extent of surgery and the need for postoperative radiation therapy. For instance, a subtotal skull base tumor resection followed by postoperative radiation therapy can yield similar outcomes as a complete resection and it can reduce the risk of operative morbidity and mortality. With modern radiation therapy technology, it is possible to achieve a very conformal radiation dose distribution with a steep radiation dose gradient beyond the target volume treated, thereby reducing the degree of collateral damage to the critical structures and normal brain parenchyma. Furthermore, most skull base tumors have sharp margins of demarcation from the surrounding normal structures and are well suited for conformal radiation techniques. Hopefully, with the sophistication of surgical and radiation techniques, patient outcomes can be further improved and toxicities can be reduced.
17.
18.
19.
20.
21.
22.
23.
24. REFERENCES 1. Shrieve DC. Basic principles of radiobiology applied to radiotherapy of benign intracranial tumors. Neurosurg Clin N Am. 2006;17(2):67–78, v. 2. Larson DA, Flickinger JC, Loeffler JS. The radiobiology of radiosurgery. Int J Radiat Oncol Biol Phys. 1993;25(3):557–561. 3. Ten Haken RK, Lawrence TS. The clinical application of intensity-modulated radiation therapy. Semin Radiat Oncol. 2006;16(4):224–231. 4. Larson DA, Coffey RJ. Radiosurgery. Devices, principles, and rationale. Clin Neurosurg. 1993;40:429–445. 5. Shrieve DC, Kooy HM, Tarbell NJ, et al. Fractionated stereotactic radiotherapy. Important Adv Oncol. 1996:205–224. 6. Hug EB. Review of skull base chordomas: Prognostic factors and long-term results of proton-beam radiotherapy. Neurosurg Focus. 2001;10(3):E11. 7. Prasad D. Clinical results of conformal radiotherapy and radiosurgery for pituitary adenoma. Neurosurg Clin N Am. 2006;17(2):129–141, vi. 8. Flickinger JC, Nelson PB, Martinez AJ, et al. Radiotherapy of nonfunctional adenomas of the pituitary gland. Results with long-term follow-up. Cancer. 1989;63(12):2409–2414. 9. Grigsby PW, Simpson JR, Fineberg B. Late regrowth of pituitary adenomas after irradiation and/or surgery. Hazard function analysis. Cancer. 1989;63(7):1308–1312. 10. Grigsby PW, Simpson JR, Stokes S, et al. Results of surgery and irradiation or irradiation alone for pituitary adenomas. J Neurooncol. 1988;6(2):129–134. 11. Tsang RW, Brierley JD, Panzarella T, et al. Radiation therapy for pituitary adenoma: Treatment outcome and prognostic factors. Int J Radiat Oncol Biol Phys. 1994;30(3):557–565. 12. McCord MW, Buatti JM, Fennell EM, et al. Radiotherapy for pituitary adenoma: Long-term outcome and sequelae. Int J Radiat Oncol Biol Phys. 1997;39(2):437–444. 13. Brada M, Rajan B, Traish D, et al. The long-term efficacy of conservative surgery and radiotherapy in the control of pituitary adenomas. Clin Endocrinol (Oxf). 1993;38(6):571–578. 14. Eastman RC, Gorden P, Glatstein E, et al. Radiation therapy of acromegaly. Endocrinol Metab Clin North Am. 1992;21(3):693– 712. 15. Ludecke DK, Lutz BS, Niedworok G. The choice of treatment after incomplete adenomectomy in acromegaly: Proton– versus high voltage radiation. Acta Neurochir (Wien). 1989;96(1–2):32– 38. 16. Jenkins PJ, Bates P, Carson MN, et al. Conventional pituitary irradiation is effective in lowering serum growth hormone and
25.
26.
27.
28.
29.
30. 31.
32.
33.
34.
35.
36.
37.
38.
insulin-like growth factor-I in patients with acromegaly. J Clin Endocrinol Metab. 2006;91(4):1239–1245. Biermasz NR, van Dulken H, Roelfsema F. Long-term followup results of postoperative radiotherapy in 36 patients with acromegaly. J Clin Endocrinol Metab. 2000;85(7):2476–2482. Grigsby PW, Simpson JR, Emami BN, et al. Prognostic factors and results of surgery and postoperative irradiation in the management of pituitary adenomas. Int J Radiat Oncol Biol Phys. 1989;16(6):1411–1417. Grigsby PW, Stokes S, Marks JE, et al. Prognostic factors and results of radiotherapy alone in the management of pituitary adenomas. Int J Radiat Oncol Biol Phys. 1988;15(5):1103–1110. Littley MD, Shalet SM, Reid H, et al. The effect of external pituitary irradiation on elevated serum prolactin levels in patients with pituitary macroadenomas. Q J Med. 1991;81(296):985–998. Tsagarakis S, Grossman A, Plowman PN, et al. Megavoltage pituitary irradiation in the management of prolactinomas: Longterm follow-up. Clin Endocrinol (Oxf). 1991;34(5):399–406. Johnston DG, Hall K, Kendall-Taylor P, et al. The long-term effects of megavoltage radiotherapy as sole or combined therapy for large prolactinomas: Studies with high definition computerized tomography. Clin Endocrinol (Oxf). 1986;24(6):675–685. Clarke SD, Woo SY, Butler EB, et al. Treatment of secretory pituitary adenoma with radiation therapy. Radiology. 1993;188(3):759–763. Hughes MN, Llamas KJ, Yelland ME, et al. Pituitary adenomas: Long-term results for radiotherapy alone and post-operative radiotherapy. Int J Radiat Oncol Biol Phys. 1993;27(5):1035–1043. Rush SC, Newall J. Pituitary adenoma: The efficacy of radiotherapy as the sole treatment. Int J Radiat Oncol Biol Phys. 1989;17(1):165–169. Estrada J, Boronat M, Mielgo M, et al. The long-term outcome of pituitary irradiation after unsuccessful transsphenoidal surgery in Cushing’s disease. N Engl J Med. 1997;336(3):172–177. Vicente A, Estrada J, de la Cuerda C, et al. Results of external pituitary irradiation after unsuccessful transsphenoidal surgery in Cushing’s disease. Acta Endocrinol (Copenh). 1991;125(5):470– 474. Howlett TA, Plowman PN, Wass JA, et al. Megavoltage pituitary irradiation in the management of Cushing’s disease and Nelson’s syndrome: Long-term follow-up. Clin Endocrinol (Oxf). 1989;31(3):309–323. Littley MD, Shalet SM, Beardwell CG, et al. Long-term followup of low-dose external pituitary irradiation for Cushing’s disease. Clin Endocrinol (Oxf). 1990;33(4):445–455. Witt TC. Stereotactic radiosurgery for pituitary tumors. Neurosurg Focus. 2003;14(5):e10. Kaylie DM, Horgan MJ, Delashaw JB, et al. A meta-analysis comparing outcomes of microsurgery and gamma knife radiosurgery. Laryngoscope. 2000;110(11):1850–1856. Kondziolka D, Lunsford LD, Flickinger JC. Comparison of management options for patients with acoustic neuromas. Neurosurg Focus. 2003;14(5):e1. Pollock BE, Driscoll CL, Foote RL, et al. Patient outcomes after vestibular schwannoma management: A prospective comparison of microsurgical resection and stereotactic radiosurgery. Neurosurgery. 2006;59(1):77–85; discussion 77–85. Myrseth E, Moller P, Pedersen PH, et al. Vestibular schwannomas: Clinical results and quality of life after microsurgery or gamma knife radiosurgery. Neurosurgery. 2005;56(5):927–935; discussion 35. Flickinger JC, Barker FG II. Clinical results: Radiosurgery and radiotherapy of cranial nerve schwannomas. Neurosurg Clin N Am. 2006;17(2):121–128, vi. Kondziolka D, Lunsford LD, McLaughlin MR, et al. Long-term outcomes after radiosurgery for acoustic neuromas. N Engl J Med. 1998;339(20):1426–1433. Lunsford LD, Niranjan A, Flickinger JC, et al. Radiosurgery of vestibular schwannomas: Summary of experience in 829 cases. J Neurosurg. 2005;102(Suppl):195–199. Flickinger JC, Kondziolka D, Niranjan A, et al. Acoustic neuroma radiosurgery with marginal tumor doses of 12 to 13 Gy. Int J Radiat Oncol Biol Phys. 2004;60(1):225–230.
Chapter 10: Radiobiology and Radiation Therapy of Skull Base Tumors 39. Inoue HK. Low-dose radiosurgery for large vestibular schwannomas: Long-term results of functional preservation. J Neurosurg. 2005;102(Suppl):111–113. 40. Petit JH, Hudes RS, Chen TT, et al. Reduced-dose radiosurgery for vestibular schwannomas. Neurosurgery. 2001;49(6):1299– 1306; discussion 306–307. 41. Paek SH, Chung HT, Jeong SS, et al. Hearing preservation after gamma knife stereotactic radiosurgery of vestibular schwannoma. Cancer. 2005;104(3):580–590. 42. Regis J, Pellet W, Delsanti C, et al. Functional outcome after gamma knife surgery or microsurgery for vestibular schwannomas. J Neurosurg. 2002;97(5):1091–1100. 43. Andrews DW, Suarez O, Goldman HW, et al. Stereotactic radiosurgery and fractionated stereotactic radiotherapy for the treatment of acoustic schwannomas: Comparative observations of 125 patients treated at one institution. Int J Radiat Oncol Biol Phys. 2001;50(5):1265–1278. 44. Iwai Y, Yamanaka K, Shiotani M, et al. Radiosurgery for acoustic neuromas: Results of low-dose treatment. Neurosurgery. 2003;53(2):282–287; discussion 7–8. 45. Weber DC, Chan AW, Bussiere MR, et al. Proton beam radiosurgery for vestibular schwannoma: Tumor control and cranial nerve toxicity. Neurosurgery. 2003;53(3):577–586; discussion 86– 88. 46. Rowe JG, Radatz MW, Walton L, et al. Gamma knife stereotactic radiosurgery for unilateral acoustic neuromas. J Neurol Neurosurg Psychiatry. 2003;74(11):1536–1542. 47. Unger F, Walch C, Schrottner O, et al. Cranial nerve preservation after radiosurgery of vestibular schwannomas. Acta Neurochir Suppl. 2002;84:77–83. 48. Friedman WA, Bradshaw P, Myers A, et al. Linear accelerator radiosurgery for vestibular schwannomas. J Neurosurg. 2006;105(5):657–661. 49. Combs SE, Thilmann C, Debus J, et al. Long-term outcome of stereotactic radiosurgery (SRS) in patients with acoustic neuromas. Int J Radiat Oncol Biol Phys. 2006;64(5):1341–1347. 50. Combs SE, Volk S, Schulz-Ertner D, et al. Management of acoustic neuromas with fractionated stereotactic radiotherapy (FSRT): Long-term results in 106 patients treated in a single institution. Int J Radiat Oncol Biol Phys. 2005;63(1):75–81. 51. Williams JA. Fractionated stereotactic radiotherapy for acoustic neuromas. Acta Neurochir (Wien). 2002;144(12):1249–1254; discussion 54. 52. Williams JA. Fractionated stereotactic radiotherapy for acoustic neuromas. Stereotact Funct Neurosurg. 2002;78(1):17–28. 53. Williams JA. Fractionated stereotactic radiotherapy for acoustic neuromas. Int J Radiat Oncol Biol Phys. 2002;54(2):500–504. 54. Williams JA. Fractionated stereotactic radiotherapy for acoustic neuromas: Preservation of function versus size. J Clin Neurosci. 2003;10(1):48–52. 55. Song DY, Williams JA. Fractionated stereotactic radiosurgery for treatment of acoustic neuromas. Stereotact Funct Neurosurg. 1999;73(1–4):45–49. 56. Shirato H, Sakamoto T, Takeichi N, et al. Fractionated stereotactic radiotherapy for vestibular schwannoma (VS): Comparison between cystic-type and solid-type VS. Int J Radiat Oncol Biol Phys. 2000;48(5):1395–1401. 57. Selch MT, Pedroso A, Lee SP, et al. Stereotactic radiotherapy for the treatment of acoustic neuromas. J Neurosurg. 2004;101(Suppl 3):362–372. 58. Harsh GR, Thornton AF, Chapman PH, et al. Proton beam stereotactic radiosurgery of vestibular schwannomas. Int J Radiat Oncol Biol Phys. 2002;54(1):35–44. 59. Chan AW, Black P, Ojemann RG, et al. Stereotactic radiotherapy for vestibular schwannomas: Favorable outcome with minimal toxicity. Neurosurgery. 2005;57(1):60–70; discussion 60–70. 60. Chung HT, Ma R, Toyota B, et al. Audiologic and treatment outcomes after linear accelerator-based stereotactic irradiation for acoustic neuroma. Int J Radiat Oncol Biol Phys. 2004;59(4):1116– 1121. 61. Bush DA, McAllister CJ, Loredo LN, et al. Fractionated proton beam radiotherapy for acoustic neuroma. Neurosurgery. 2002;50(2):270–273; discussion 3–5.
171
62. Meijer OW, Vandertop WP, Baayen JC, et al. Single-fraction vs. fractionated linac-based stereotactic radiosurgery for vestibular schwannoma: A single-institution study. Int J Radiat Oncol Biol Phys. 2003;56(5):1390–1396. 63. Chang SD, Gibbs IC, Sakamoto GT, et al. Staged stereotactic irradiation for acoustic neuroma. Neurosurgery. 2005;56(6):1254– 1261; discussion 61–63. 64. Goldsmith B, McDermott MW. Meningioma. Neurosurg Clin N Am. 2006;17(2):111–120, vi. 65. Glaholm J, Bloom HJ, Crow JH. The role of radiotherapy in the management of intracranial meningiomas: The Royal Marsden Hospital experience with 186 patients. Int J Radiat Oncol Biol Phys. 1990;18(4):755–761. 66. Soyuer S, Chang EL, Selek U, et al. Radiotherapy after surgery for benign cerebral meningioma. Radiother Oncol. 2004;71(1):85–90. 67. Uy NW, Woo SY, Teh BS, et al. Intensity-modulated radiation therapy (IMRT) for meningioma. Int J Radiat Oncol Biol Phys. 2002;53(5):1265–1270. 68. Pirzkall A, Debus J, Haering P, et al. Intensity modulated radiotherapy (IMRT) for recurrent, residual, or untreated skull-base meningiomas: Preliminary clinical experience. Int J Radiat Oncol Biol Phys. 2003;55(2):362–372. 69. Sajja R, Barnett GH, Lee SY, et al. Intensity-modulated radiation therapy (IMRT) for newly diagnosed and recurrent intracranial meningiomas: Preliminary results. Technol Cancer Res Treat. 2005;4(6):675–682. 70. Pollock BE, Stafford SL. Results of stereotactic radiosurgery for patients with imaging defined cavernous sinus meningiomas. Int J Radiat Oncol Biol Phys. 2005;62(5):1427–1431. 71. Spiegelmann R, Nissim O, Menhel J, et al. Linear accelerator radiosurgery for meningiomas in and around the cavernous sinus. Neurosurgery. 2002;51(6):1373–1379; discussion 9–80. 72. Lee JY, Niranjan A, McInerney J, et al. Stereotactic radiosurgery providing long-term tumor control of cavernous sinus meningiomas. J Neurosurg. 2002;97(1):65–72. 73. Roche PH, Pellet W, Fuentes S, et al. Gamma knife radiosurgical management of petroclival meningiomas results and indications. Acta Neurochir (Wien). 2003;145(10):883–888; discussion 8. 74. Roche PH, Regis J, Dufour H, et al. Gamma knife radiosurgery in the management of cavernous sinus meningiomas. J Neurosurg. 2000;93(Suppl 3):68–73. 75. Subach BR, Lunsford LD, Kondziolka D, et al. Management of petroclival meningiomas by stereotactic radiosurgery. Neurosurgery. 1998;42(3):437–443; discussion 43–45. 76. Jalali R, Loughrey C, Baumert B, et al. High precision focused irradiation in the form of fractionated stereotactic conformal radiotherapy (SCRT) for benign meningiomas predominantly in the skull base location. Clin Oncol (R Coll Radiol). 2002;14(2):103–109. 77. Alheit H, Saran FH, Warrington AP, et al. Stereotactically guided conformal radiotherapy for meningiomas. Radiother Oncol. 1999;50(2):145–150. 78. Debus J, Wuendrich M, Pirzkall A, et al. High efficacy of fractionated stereotactic radiotherapy of large base-of-skull meningiomas: Long-term results. J Clin Oncol. 2001;19(15):3547– 3553. 79. Milker-Zabel S, Zabel-du Bois A, Huber P, et al. Fractionated stereotactic radiation therapy in the management of benign cavernous sinus meningiomas: Long-term experience and review of the literature. Strahlenther Onkol. 2006;182(11):635–640. 80. Selch MT, Ahn E, Laskari A, et al. Stereotactic radiotherapy for treatment of cavernous sinus meningiomas. Int J Radiat Oncol Biol Phys. 2004;59(1):101–111. 81. Brell M, Villa S, Teixidor P, et al. Fractionated stereotactic radiotherapy in the treatment of exclusive cavernous sinus meningioma: Functional outcome, local control, and tolerance. Surg Neurol. 2006;65(1):28–33; discussion 4. 82. Noel G, Bollet MA, Calugaru V, et al. Functional outcome of patients with benign meningioma treated by 3D conformal irradiation with a combination of photons and protons. Int J Radiat Oncol Biol Phys. 2005;62(5):1412–1422.
172
Lo et al.
83. Vernimmen FJ, Harris JK, Wilson JA, et al. Stereotactic proton beam therapy of skull base meningiomas. Int J Radiat Oncol Biol Phys. 2001;49(1):99–105. 84. Wenkel E, Thornton AF, Finkelstein D, et al. Benign meningioma: Partially resected, biopsied, and recurrent intracranial tumors treated with combined proton and photon radiotherapy. Int J Radiat Oncol Biol Phys. 2000;48(5):1363–1370. 85. Noel G, Habrand JL, Mammar H, et al. Highly conformal therapy using proton component in the management of meningiomas. Preliminary experience of the Centre de Protontherapie d’Orsay. Strahlenther Onkol. 2002;178(9):480–485. 86. Weber DC, Lomax AJ, Rutz HP, et al. Spot-scanning proton radiation therapy for recurrent, residual or untreated intracranial meningiomas. Radiother Oncol. 2004;71(3):251–258. 87. Suh JH, Gupta N. Role of radiation therapy and radiosurgery in the management of craniopharyngiomas. Neurosurg Clin N Am. 2006;17(2):143–148, vi-vii. 88. Luu QT, Loredo LN, Archambeau JO, et al. Fractionated proton radiation treatment for pediatric craniopharyngioma: Preliminary report. Cancer J. 2006;12(2):155–159. 89. Ulfarsson E, Lindquist C, Roberts M, et al. Gamma knife radiosurgery for craniopharyngiomas: Long-term results in the first Swedish patients. J Neurosurg. 2002;97(5)(Suppl):613–622. 90. Schulz-Ertner D, Frank C, Herfarth KK, et al. Fractionated stereotactic radiotherapy for craniopharyngiomas. Int J Radiat Oncol Biol Phys. 2002;54(4):1114–1120. 91. Selch MT, DeSalles AA, Wade M, et al. Initial clinical results of stereotactic radiotherapy for the treatment of craniopharyngiomas. Technol Cancer Res Treat. 2002;1(1):51–59. 92. Kalapurakal JA, Kepka A, Bista T, et al. Fractionated stereotactic radiotherapy for pediatric brain tumors: The Chicago children’s experience. Childs Nerv Syst. 2000;16(5):296–302; discussion 3. 93. Kortmann RD, Becker G, Perelmouter J, et al. Geometric accuracy of field alignment in fractionated stereotactic conformal radiotherapy of brain tumors. Int J Radiat Oncol Biol Phys. 1999;43(4):921–926. 94. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. Management of cystic craniopharyngiomas with phosphorus-32 intracavitary irradiation. Neurosurgery. 2004;54(4):813–820; discussion 20–22. 95. Voges J, Sturm V, Lehrke R, et al. Cystic craniopharyngioma: Long-term results after intracavitary irradiation with stereotactically applied colloidal β-emitting radioactive sources. Neurosurgery. 1997;40(2):263–269; discussion 9–70. 96. Powell S, Peters N, Harmer C. Chemodectoma of the head and neck: Results of treatment in 84 patients. Int J Radiat Oncol Biol Phys. 1992;22(5):919–924. 97. Wang ML, Hussey DH, Doornbos JF, et al. Chemodectoma of the temporal bone: A comparison of surgical and radiotherapeutic results. Int J Radiat Oncol Biol Phys. 1988;14(4):643–648. 98. Hinerman RW, Mendenhall WM, Amdur RJ, et al. Definitive radiotherapy in the management of chemodectomas arising in the temporal bone, carotid body, and glomus vagale. Head Neck. 2001;23(5):363–371. 99. Konefal JB, Pilepich MV, Spector GJ, et al. Radiation therapy in the treatment of chemodectomas. Laryngoscope. 1987;97(11):1331–1335. 100. Sharma PD, Johnson AP, Whitton AC. Radiotherapy for jugulotympanic paragangliomas (glomus jugulare tumours). J Laryngol Otol. 1984;98(6):621–629. 101. Cummings BJ, Beale FA, Garrett PG, et al. The treatment of glomus tumors in the temporal bone by megavoltage radiation. Cancer. 1984;53(12):2635–2640. 102. Pemberton LS, Swindell R, Sykes AJ. Radical radiotherapy alone for glomus jugulare and tympanicum tumours. Oncol Rep. 2005;14(6):1631–1633. 103. Krych AJ, Foote RL, Brown PD, et al. Long-term results of irradiation for paraganglioma. Int J Radiat Oncol Biol Phys. 2006;65(4):1063–1066. 104. Springate SC, Weichselbaum RR. Radiation or surgery for chemodectoma of the temporal bone: A review of local control and complications. Head Neck. 1990;12(4):303–307.
105. Liscak R, Vladyka V, Wowra B, et al. Gamma Knife radiosurgery of the glomus jugulare tumour—early multicentre experience. Acta Neurochir (Wien). 1999;141(11):1141–1146. 106. Foote RL, Pollock BE, Gorman DA, et al. Glomus jugulare tumor: Tumor control and complications after stereotactic radiosurgery. Head Neck. 2002;24(4):332–338; discussion 8–9. 107. Eustacchio S, Leber K, Trummer M, et al. Gamma knife radiosurgery for glomus jugulare tumours. Acta Neurochir (Wien). 1999;141(8):811–818. 108. Liscak R, Vladyka V, Simonova G, et al. Leksell gamma knife radiosurgery of the tumor glomus jugulare and tympanicum. Stereotact Funct Neurosurg. 1998;70(Suppl 1):152–160. 109. Jordan JA, Roland PS, McManus C, et al. Stereotactic radiosurgery for glomus jugulare tumors. Laryngoscope. 2000;110(1):35–38. 110. Gerosa M, Visca A, Rizzo P, et al. Glomus jugulare tumors: The option of gamma knife radiosurgery. Neurosurgery. 2006;59(3):561–569; discussion 9. 111. Sheehan J, Kondziolka D, Flickinger J, et al. Gamma knife surgery for glomus jugulare tumors: An intermediate report on efficacy and safety. J Neurosurg. 2005;102(Suppl):241– 246. 112. Krishnan S, Foote RL, Brown PD, et al. Radiosurgery for cranial base chordomas and chondrosarcomas. Neurosurgery. 2005;56(4):777–784; discussion 84. 113. Kondziolka D, Lunsford LD, Flickinger JC. The role of radiosurgery in the management of chordoma and chondrosarcoma of the cranial base. Neurosurgery. 1991;29(1):38–45; discussion 6. 114. Muthukumar N, Kondziolka D, Lunsford LD, et al. Stereotactic radiosurgery for chordoma and chondrosarcoma: Further experiences. Int J Radiat Oncol Biol Phys. 1998;41(2):387–392. 115. Debus J, Schulz-Ertner D, Schad L, et al. Stereotactic fractionated radiotherapy for chordomas and chondrosarcomas of the skull base. Int J Radiat Oncol Biol Phys. 2000;47(3):591–596. 116. Noel G, Feuvret L, Dhermain F, et al. Chordomas of the base of the skull and upper cervical spine. 100 patients irradiated by a 3D conformal technique combining photon and proton beams. Cancer Radiother. 2005;9(3):161–174. 117. Noel G, Habrand JL, Mammar H, et al. Combination of photon and proton radiation therapy for chordomas and chondrosarcomas of the skull base: The Centre de Protontherapie D’Orsay experience. Int J Radiat Oncol Biol Phys. 2001;51(2):392– 398. 118. Noel G, Habrand JL, Jauffret E, et al. Radiation therapy for chordoma and chondrosarcoma of the skull base and the cervical spine. Prognostic factors and patterns of failure. Strahlenther Onkol. 2003;179(4):241–248. 119. Weber DC, Rutz HP, Pedroni ES, et al. Results of spot-scanning proton radiation therapy for chordoma and chondrosarcoma of the skull base: The Paul Scherrer Institut experience. Int J Radiat Oncol Biol Phys. 2005;63(2):401–409. 120. Hug EB, Slater JD. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. Neurosurg Clin N Am. 2000;11(4):627–638. 121. Hug EB, Loredo LN, Slater JD, et al. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. J Neurosurg. 1999;91(3):432–439. 122. Benk V, Liebsch NJ, Munzenrider JE, et al. Base of skull and cervical spine chordomas in children treated by high-dose irradiation. Int J Radiat Oncol Biol Phys. 1995;31(3):577–581. 123. Schulz-Ertner D, Nikoghosyan A, Thilmann C, et al. Results of carbon ion radiotherapy in 152 patients. Int J Radiat Oncol Biol Phys. 2004;58(2):631–640. 124. Schulz-Ertner D, Nikoghosyan A, Didinger B, et al. Carbon ion radiation therapy for chordomas and low grade chondrosarcomas—current status of the clinical trials at GSI. Radiother Oncol. 2004;73(Suppl 2):S53–S56. 125. Rosenberg AE, Nielsen GP, Keel SB, et al. Chondrosarcoma of the base of the skull: A clinicopathologic study of 200 cases with emphasis on its distinction from chordoma. Am J Surg Pathol. 1999;23(11):1370–1378.
Chapter 10: Radiobiology and Radiation Therapy of Skull Base Tumors 126. Schulz-Ertner D, Nikoghosyan A, Hof H, et al. Carbon ion radiotherapy of skull base chondrosarcomas. Int J Radiat Oncol Biol Phys. 2007;67(1):171–177. 127. Mould RF, Tai TH. Nasopharyngeal carcinoma: Treatments and outcomes in the 20th century. Br J Radiol. 2002;75(892):307– 339. 128. Baujat B, Audry H, Bourhis J, et al. Chemotherapy as an adjunct to radiotherapy in locally advanced nasopharyngeal carcinoma. Cochrane Database Syst Rev. 2006;(4):CD004329. 129. O’Meara WP, Lee N. Advances in nasopharyngeal carcinoma. Curr Opin Oncol. 2005;17(3):225–230. 130. Teo PM, Ma BB, Chan AT. Radiotherapy for nasopharyngeal carcinoma—transition from two-dimensional to threedimensional methods. Radiother Oncol. 2004;73(2):163–172. 131. Kam MK, Chau RM, Suen J, et al. Intensity-modulated radiotherapy in nasopharyngeal carcinoma: Dosimetric advantage over conventional plans and feasibility of dose escalation. Int J Radiat Oncol Biol Phys. 2003;56(1):145–157. 132. Lee N, Xia P, Quivey JM, et al. Intensity-modulated radiotherapy in the treatment of nasopharyngeal carcinoma: An update of the UCSF experience. Int J Radiat Oncol Biol Phys. 2002;53(1):12– 22. 133. Kam MK, Teo PM, Chau RM, et al. Treatment of nasopharyngeal carcinoma with intensity-modulated radiotherapy: The Hong Kong experience. Int J Radiat Oncol Biol Phys. 2004;60(5):1440– 1450. 134. Kwong DL, Pow EH, Sham JS, et al. Intensity-modulated radiotherapy for early-stage nasopharyngeal carcinoma: A prospective study on disease control and preservation of salivary function. Cancer. 2004;101(7):1584–1593. 135. Parsons JT, Kimsey FC, Mendenhall WM, et al. Radiation therapy for sinus malignancies. Otolaryngol Clin North Am. 1995;28(6):1259–1268.
173
136. Blanco AI, Chao KS, Ozyigit G, et al. Carcinoma of paranasal sinuses: Long-term outcomes with radiotherapy. Int J Radiat Oncol Biol Phys. 2004;59(1):51–58. 137. Mendenhall WM, Mendenhall CM, Riggs CE Jr, et al. Sinonasal undifferentiated carcinoma. Am J Clin Oncol. 2006;29(1):27–31. 138. Eich HT, Staar S, Micke O, et al. Radiotherapy of esthesioneuroblastoma. Int J Radiat Oncol Biol Phys. 2001;49(1):155–160. 139. Monroe AT, Hinerman RW, Amdur RJ, et al. Radiation therapy for esthesioneuroblastoma: Rationale for elective neck irradiation. Head Neck. 2003;25(7):529–534. 140. Chao KS, Kaplan C, Simpson JR, et al. Esthesioneuroblastoma: The impact of treatment modality. Head Neck. 2001;23(9):749– 757. 141. Gruber G, Laedrach K, Baumert B, et al. Esthesioneuroblastoma: Irradiation alone and surgery alone are not enough. Int J Radiat Oncol Biol Phys. 2002;54(2):486–491. 142. Loy AH, Reibel JF, Read PW, et al. Esthesioneuroblastoma: Continued follow-up of a single institution’s experience. Arch Otolaryngol Head Neck Surg. 2006;132(2):134–138. 143. Diaz EM Jr, Johnigan RH III, Pero C, et al. Olfactory neuroblastoma: The 22-year experience at one comprehensive cancer center. Head Neck. 2005;27(2):138–149. 144. Combs SE, Konkel S, Schulz-Ertner D, et al. Intensity modulated radiotherapy (IMRT) in patients with carcinomas of the paranasal sinuses: Clinical benefit for complex shaped target volumes. Radiat Oncol. 2006;1:23. 145. Pacholke HD, Amdur RJ, Louis DA, et al. The role of intensity modulated radiation therapy for favorable stage tumor of the nasal cavity or ethmoid sinus. Am J Clin Oncol. 2005;28(5):474– 478. 146. Duthoy W, Boterberg T, Claus F, et al. Postoperative intensitymodulated radiotherapy in sinonasal carcinoma: Clinical results in 39 patients. Cancer. 2005;104(1):71–82.
11 Chemotherapy for Skull Base Tumors Bilal Ahmed, Ehab Y. Hanna, and Merrill S. Kies
control and organ preservation. We will review data largely generated in phase II studies, in an attempt to define active systemic regimens with acceptable toxicity and applicability to combined modality programs. In doing so, we should also consider directions for future clinical investigations.
CHEMOTHERAPY FOR SKULL BASE TUMORS Skull base tumors are rare, and the site of origin may be difficult to determine as often there is involvement of multiple skull base compartments with variable effects on important neurovascular structures (1). Infiltrating skull base cancers may arise from the nasopharynx, paranasal sinuses, major and minor salivary glands, and skin. Treatment of nasopharyngeal cancers has been the subject of excellent recent reviews (2,3), and will not be addressed in this chapter. We will focus our discussion on chemotherapy for sinonasal cancers since they are a common malignancy that involve or invade the cranial base. Sinonasal cancers are uncommon and constitute 3% of head and neck cancers (4). These cancers present a special set of problems for clinicians as diagnosis is often delayed and symptomatology tends to emerge with involvement of vital local structures. Infiltrating disease in the nasal cavity may precipitate epistaxis or obstruction; orbital involvement may produce proptosis or diplopia; and skull base and intracranial tumor leads to variable neurologic deficits. Squamous cell carcinoma, adenocarcinomas, adenoid cystic carcinoma, neuroendocrine cancers, melanomas, Merkel cell malignancies, sarcomas, and lymphomas are among a diverse group of malignancies encountered. Surgery has generally been the mainstay of therapy for these tumors but frequently resections are subtotal necessitating multidisciplinary management. Even radical surgical excisions such as craniofacial resection, total maxillectomy, or orbital exenteration produce in aggregate less than 50% 5-year survival with both local and distant tumor failure patterns common (3). So with time, the skull base team has evolved to include surgical, radiation, and medical oncologists; neuro-otologists; oncologic nursing; rehabilitation specialists; social service staff; and psychology consultants. Proximity to many vital structures renders surgical resection often complicated. However, for squamous cell and neuroendocrine malignancies, surgery has traditionally been the cornerstone of management with radiation therapy administered as a postoperative adjunct for patients with locally advanced disease. Chemotherapy has previously been reserved for palliation in patients with incurable, locally far advanced, or metastatic disease. Standard surgical approaches may require orbitectomy or rhinectomy in some patients, obviously major debilitating events. So, as we consider that tumor control and health-related quality of life (5) are important outcome measures in assessing the potential benefits of combined modality therapy for skull base cancers, the subject of this chapter will be on the emerging role of systemic drug therapy as a component of primary treatment plans for selected patients. Induction chemotherapy, described as neoadjuvant therapy in some centers, may be of particular value in therapeutic strategies designed with dual objectives of tumor
TREATMENT RESULTS FOR PARANASAL SINUS CANCER Squamous Cell Carcinoma For patients with locally advanced disease, surgical resection followed by postoperative radiotherapy is common practice. Nonetheless, frequent involvement of vital structures and high locoregional failure risks call for a revision of older treatment plans. Overall disease-free survival at 5 years is less than 50% and 25% to 30% with more advanced disease (3,6,7). That is why new approaches such as induction chemotherapy followed by concomitant chemoradiotherapy or surgical resection with postoperative radiotherapy have been an area of active interest in recent studies. A Swedish group (8) presented an analysis of a mix of patients with paranasal sinus cancers treated with induction cisplatin and fluorouracil before radiotherapy then surgery. Seventy percent achieved a clinical partial tumor response to the drug therapy. Lee et al. (9) reported a University of Chicago experience in 16 patients receiving induction chemotherapy with cisplatin and infusional fluorouracil, then surgical resection and postoperative chemoradiotherapy with concomitant fluorouracil, hydroxyurea, and radiotherapy. Ten-year local and distant disease control exceeded 90%. For early-stage disease without extensive locoregional involvement, excision of the tumor followed by radiation seems to be the reasonable approach. However, for advanced stage M0 disease we recommend that trimodal strategy (surgery, radiation, and chemotherapy) should be used. The optimal sequence for chemotherapy and surgery or radiotherapy is not very clear. Isobe et al. reported a large, retrospective analysis examining the effect of sequencing chemotherapy in the management of squamous cell cancer of the maxillary sinus. Patients were divided into two groups, those who received cisplatin and peplomycin in the neoadjuvant settings and those who received concurrent chemoradiation. The study failed to show a significant effect for treatment sequence on survival (10). Intra-arterial (IA) chemotherapy is an alternative approach. This strategy aims to induce tumor regression to limit surgery and the associated detrimental effects on quality of life. Paranasal sinus cancers supplied by the internal maxillary artery are the most suitable targets for this highly selective approach. IA chemotherapy is facilitated greatly by technologic improvements and has the potential for increasing drug concentration at tumor sites and lowering systemic exposure. Increased tumor drug concentration for cisplatin is achievable and may be important, because 175
176
Ahmed et al.
dose-response relations seem to exist in head and neck cancers. Lee et al. (11) reported that IA cisplatin could be administered through superselective indwelling catheters and combined with systemic chemotherapy to achieve an overall response rate of 80% (30% complete) in advanced paranasal sinus tumors. The MD Anderson Cancer Center has further explored this approach with induction chemotherapy consisting of IA cisplatin and systemic paclitaxel and ifosfamide (12). To be eligible, disease was locally advanced and would require orbital exenteration or craniofacial resection. If substantial tumor response obtained, patients proceeded directly to definitive local treatment with radiotherapy, and surgery was avoided. Initial results with this approach were promising in terms of tumor response and local control. Eleven of 19 evaluable patients achieved a tumor response, and eyesparing surgery followed by radiotherapy was feasible in seven patients and radiotherapy without surgery in seven patients. At completion of treatment, 14 patients (61%) were disease-free with organ-preservation in 21 of 24 (88%). Overall survival at 2 years was 60%. Toxicity was substantial with two patients experiencing cerebral vascular ischemia (one transient) and three cranial neuropathy (one not reversible). Madison and colleagues have also reported promising results, from the University of Tennessee, and a series of 11 patients treated with preoperative, IA cisplatin and concurrent radiation therapy followed by a conservative craniofacial resection (13). In our experience with a mix of tumor histologies, progression-free survival at 5 years is 67%. All agree that further study in a research setting is warranted. Efforts are now focused on improved assessment of the response to therapy, and integration of targeted systemic drug administration with radiotherapy and surgery. For locally extensive disease, where a complete excision is likely to result in significant morbidity, we presently consider use of induction chemotherapy with docetaxel, cisplatin, and fluorouracil to induce disease regression before definitive local therapy, surgery, or possibly concomitant chemoradiotherapy. This strategy can be expected to reduce risk of distant disease recurrence; if there is a marked tumor response, there may also be an enhancement of the efficacy of local therapy. Cisplatin and infusional fluorouracil is a highly active drug regimen first developed as an induction regimen (14,15). Since then, the value of chemotherapy for the primary treatment of locally advanced squamous cancers of the pharynx and larynx has been demonstrated in a series of controlled trials (14,16–19) and meta-analyses (20). Concomitant chemotherapy and radiotherapy is superior to radiotherapy administered as a single modality with improved local tumor control and overall survival. Cisplatin, administered 100 mg/m2 during weeks 1, 4, and 7 with fractionated radiotherapy once daily now represents an accepted standard of care for patients with stage III and IV M0 disease. Study of concomitant cisplatin and radiotherapy in the postoperative setting has also shown a survival advantage for high-risk patients with positive resection margins or demonstrated extracapsular tumor extension (21,22,23). Moreover, alternative drug and radiotherapy schedules are under study. The updated meta-analysis of Monnerat and colleagues (24) also describes a survival advantage for patients receiving cisplatin and fluorouracil as the induction schedule. However, induction chemotherapy has not yet been accepted as a standard of care and there are ongoing randomized trials (14) addressing the value of this approach. Renewed interest in the administration of induction chemotherapy for squamous cancers of multisite origin has followed the reports of increased chemotherapy efficacy with
the addition of a taxane, such as paclitaxel or docetaxel, to the basic cisplatin and fluorouracil combination. Hitt et al. (25) compared induction programs with cisplatin and fluorouracil with or without paclitaxel. They observed a significantly higher response rate with the experimental regimen with no overall increase in toxicity. Patients with disease considered unresectable at diagnosis then proceeded to concomitant chemoradiotherapy and there was a survival advantage for the three-drug arm (p = 0.04). More recent presentations by Vermorken (26) and Posner (27) have furthered research interest in this area. In an EORTC phase III study comparing induction cisplatin and fluorouracil with docetaxel, cisplatin, and fluorouracil, 358 patients were entered. Following induction chemotherapy, all patients received radiotherapy as a single local treatment modality. With median follow-up of 32 months, the three-drug arm achieved superior tumor responses with significant progression-free and overall survival hazard ratios, 0.72 and 0.73, respectively. Also, more toxic deaths 5.5% versus 2.3% were observed in the cisplatin and fluorouracil group. In TAX 324, patients were randomized to receive induction chemotherapy with docetaxel 75 mg/m2 , cisplatin 100 mg/m2 , and fluorouracil 1000 mg/m2 /d (TPF) in a continuous infusion over 96 hours versus the standard cisplatin and fluorouracil schedule. Three treatment cycles were administered before patients proceeded to concomitant chemoradiotherapy with carboplatin AUC 1.5 administered weekly. Surgery was performed as salvage for residual disease. There appears to be a significant overall survival advantage for the patients receiving TPF. With a hazard ratio 0.70, 62% of patients receiving TPF were alive at 36 months versus 48%. Moreover, the authors have reported that there is the potential for TPF to have had a favorable effect on local disease control. At the MD Anderson Cancer Center, we are commencing a phase II trial with induction TPF followed by definitive chemoradiation for responding patients and surgical salvage if there is evidence for residual disease after the conclusion of radiotherapy. Patients whose tumors failed to respond to the induction program will proceed to definitive surgical resection and postoperative radiotherapy.
Sinonasal Undifferentiated Carcinoma Sinonasal undifferentiated carcinoma is a rare, highly aggressive, and distinct entity believed to have originated from the Schneiderian epithelium or from the nasal ectoderm of the paranasal sinuses. These cancers often present as a rapidly enlarging mass, involving multiple sites, and extending beyond the anatomic confines of the sinonasal tract. Associated symptoms reflect an enlarging mass effect, nasal obstruction, epistaxis, proptosis, facial pain, and cranial nerve palsies (28,29). This tumor is highly virulent. Frierson and colleagues (28) reported a median survival of 12.4 months in patients treated with radiotherapy, with or without chemotherapy. Only a single patient in this group underwent craniofacial resection. At the University of Virginia, experimental induction chemoradiotherapy with cyclophosphamide, doxorubicin, and vincristine is followed by craniofacial resection with reports of 65% 2-year survival in patients completing the treatment sequence. Subsequently, Rischin et al. (30) reported a series of seven patients treated with induction chemotherapy with cisplatin or carboplatin and infusional fluorouracil followed by radiotherapy with concomitant platinum administered during the first and last weeks of radiation. Four patients were disease free at 8 to 51 months follow-up. Kim and
Chapter 11: Chemotherapy for Skull Base Tumors Table 1 Kadish Staging System Stage
Tumor extension
A B C
Tumor limited to the nasal cavity Tumor involves nasal cavity and paranasal sinuses Tumor extends beyond the nasal cavity and paranasal sinuses
Enepekides (31) have reviewed retrospective reports from the literature. Prospective studies are needed to better identify molecular correlates that may predict for more effective treatment sequence, and a genomic or biomarker signature that can be used to select patients for more targeted and individualized medical therapy.
Esthesioneuroblastoma Esthesioneuroblastoma is a rare malignancy that originates from the olfactory epithelium. It comprises approximately 6% of paranasal tumors and less than 1% of all malignancies (32). Most often, this cancer arises from the superior nasal vault and invades the paranasal sinuses, orbit, and skull base, extending to CNS. Distant metastases occur in 20% to 25% of patients, with brain, lung, and bone as common sites. The Kadish staging system is presented in Table 1. Traditionally, this tumor has been treated with craniofacial resection followed by radiation therapy, and this approach is effective in Kadish stage A and B lesions, with long-term control exceeding to 75%. Chemotherapy has generally been reserved for recurrent or metastatic disease not amenable to curative local therapy. However, some more recent evidence is that chemotherapy should be integrated into primary management of advanced tumors. Resto et al. (33) have reported a retrospective review of 20 patients treated at Johns Hopkins University Hospital. These patients underwent surgical resection; with median follow up of 3.9 years, 12 patients were recurrence free. Nine of the 12 had negative surgical margins. All but two patients also received postoperative radiotherapy. Three patients had received induction chemotherapy with doxorubicin-based regimens with no response observed. Two patients received chemotherapy in sequence with radiation. A tumor response was observed to an alternating regimen with cyclophosphamide, doxorubicin, and vincristine (CAV) and cisplatin and etoposide. Also, one patient received postoperative cisplatin and etoposide with radiation, and was free of tumor recurrence at the time of the report. Polin et al. (34) analyzed 34 consecutive patients treated at the University of Virginia. In a multivariate regression analysis of potential prognostic factors, 21 patients were identified who had preoperative radiation or chemotherapy (with alkylatorbased regimens). Responding patients (n = 14) had improved survival. Overall, 5- and 10-year survival rates were 81% and 55%, respectively. Further, this group favors inclusion of chemotherapy as a component of multimodal care for patients with advanced disease. McLean et al. (35), from Emory University, have recently described their institutional experience with a total of 21 patients. Median follow up was 47 months. Surgical resection was performed in 90% of patients and postoperative radiotherapy in 73%. Multimodality therapy with chemotherapy was administered to seven patients. Overall, eight patients (38%) have had local disease recurrence, and the 5-year overall survival was 58%. The authors concluded that there should be further investigation of multimodal therapy with craniofacial resection, intensity modulated radiation therapy, and chemotherapy for patients with Kadish C lesions.
177
Adenoid Cystic Carcinoma Adenoid cystic carcinoma (ACC) originates from the salivary, lacrimal, and other exocrine glands. There are differences in gene and protein expression between adenoid cystic cancers and normal salivary glands, with over expression of c-kit (36,37). This is the most common malignancy of minor salivary glands. Skull base involvement is frequent (38), a presumed consequence of the tendency for neurotropic tumor progression, often with “skip” metastases. There may be tumor progression along the V2 and V3 branches of the trigeminal nerve to the skull base with resultant facial numbness. Tumor involvement of the skull base by direct extension also may obtain by direct invasion through the paranasal sinuses. Esmali et al. recently reported a case series demonstrating the aggressive nature of this tumor where all seven patients had perineural and skull invasion and had to be treated with orbital exenteration (39). ACCs are not highly sensitive to cytotoxic chemotherapy although regimens consisting of cyclophosphamide, doxorubicin, and cisplatin (CAP) or vinorelbine and cisplatin will induce disease regression in subsets of patients with metastatic disease (40). Surgery is clearly the mainstay for the treatment of localized disease, however, given the high incidence of perineural invasion, postoperative radiation therapy is generally recommended to encompass associated nervous pathways, up to and including the base of the skull. Notably, a significant fraction of patients may present with locally advanced and unresectable disease, e.g., in the nasopharynx. Definitive concomitant chemoradiotherapy with weekly paclitaxel and carboplatin (41) or cisplatin administered during weeks 1, 4, and 7 of a daily radiotherapy treatment schedule (42) may have substantial activity with sustained disease remission. Also to be considered is that ACC may often have a prolonged clinical course with local recurrences and distant metastases occurring years after presentation (43,44). For the asymptomatic patient, particularly if older, with very slowprogressing lung metastases a period of observation may be appropriate, especially if comorbidities are more immediately problematic. However, for disease that is more obviously advanced, not amenable to local therapy, and objectively progressive often with lung and bone involvement, systemic chemotherapy or investigational treatment should be considered. CAP has been reported to induce tumor responses in as high as 46% patients (45). Platins, fluorouracil, doxorubicin, and vinorelbine are recognized as having some efficacy as single agents (46). In an early report, Glisson et al. (47) observed stabilization of metastasis in 60% of patients treated with gefitinib, the oral epidermal growth factor receptor tyrosine kinase inhibitor, receiving 250 mg daily. Hotte et al. reported no objective responses in 15 patients with overexpression of wild-type C-kit (48). Molecular markers predictive of activity for specific drug strategies have not yet been found effective in ACC.
Neuroendocrine Carcinoma Neuroendocrine carcinomas are rare sinonasal malignancies. First described in 1982 (49), this may be an aggressive cancer with poorly differentiated histology, especially the small cell subtype, with high risk for locoregional and distant disease recurrence and overall poor survival (50). Chemotherapy is under investigation as a component of primary management. At the Massachussetts General Hospital, 19 patients with either neuroblastoma or neuroendocrine tumors received two induction cycles with cisplatin and etoposide followed by
178
Ahmed et al.
proton-photon therapy. Responders then received an additional two chemotherapy cycles, with achievement of a 5-year survival rate 74% (51). A similar approach followed by Bhattacharyya et al. with induction chemotherapy consisting of cisplatin and etoposide followed by proton therapy again showed considerable activity in patients with quite bulky disease (52). As has been demonstrated in the treatment of small cell lung cancers, chemotherapy should be considered as a component of primary management in patients with locallyadvanced, poorly-differentiated disease. The experience with chemotherapy has largely been with platin and etoposide combinations, but other agents with potential activity are doxorubicin, cyclophosphamide, and camptothecins. In patients with advanced disease, we support a strategy with induction cisplatin 80 mg/m2 day 1 and etoposide 100 mg/m2 IV dose equivalent days 1–3 to be followed by concomitant chemoradiotherapy for responding patients. Certainly, an alternative may be to initiate treatment with concomitant chemoradiotherapy. Surgical resection is reserved for patients with tumors not responsive to drug or for salvage (6). It is not clear if heavy particle radiation is superior to intensitymodulated therapy.
9.
10.
11.
12. 13.
14. 15.
16.
17.
CONCLUSION In our view, there is an accumulating experience with chemotherapy as a component of primary management for many patients with skull base and sinonasal cancers. Much work needs to be done, and we strongly endorse support of clinical trials. Given the relative rarity of this group of diseases, phase II trials will be more likely and often may need to be performed in cooperative group settings. Innovative studies, with correlative scientific inquiries, tumor banking, and long-term outcome measures of tumor control, patient performance status, and life quality are imperative. For patients presenting with locally advanced disease and as we integrate systemic with locoregional therapy for patients presenting with locally advanced disease, complex treatment planning is necessary for optimal results. REFERENCES 1. Raso JL, Gusmao S. Transbasal approach to skull base tumors: Evaluation and proposal of classification. Surg Neurol. 2006;65(Suppl 1):S1:33–S1:38. 2. Tao Q, Chan AT. Nasopharyngeal carcinoma: Molecular pathogenesis and therapeutic developments. Expert Rev Mol Med. 2007;9(12):1–24. 3. Day TA, Beas RA, Schlosser RJ, et al. Management of paranasal sinus malignancy. Curr Treat Options Oncol. 2005;6(1):3–18. 4. Morita A, Sekhar LN, Wright DC. Current concepts in the management of tumors of the skull base. Cancer Control. 1998;5(2):138–149. 5. Kelleher MO, Fernandes MF, Sim DW, et al. Health-related quality of life in patients with skull base tumours. Br J Neurosurg. 2002;16(1):16–20. 6. Diaz EM Jr, Kies MS. Chemotherapy for skull base cancers. Otolaryngol Clin North Am. 2001;34(6):1079–1085, viii. 7. Blanco AI, Chao KS, Ozyigit G, et al. Carcinoma of paranasal sinuses: Long-term outcomes with radiotherapy. Int J Radiat Oncol Biol Phys. 2004;59(1):51–58. 8. Bjork-Eriksson T, Mercke C, Petruson B, et al. Potential impact on tumor control and organ preservation with cisplatin and 5-fluorouracil for patients with advanced tumors of the
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
paranasal sinuses and nasal fossa. A prospective pilot study. Cancer. 1992;70(11):2615–2620. Lee MM, Vokes EE, Rosen A, et al. Multimodality therapy in advanced paranasal sinus carcinoma: Superior long-term results. Cancer J Sci Am. 1999;5(4):219–223. Isobe K, Uno T, Hanazawa T, et al. Preoperative chemotherapy and radiation therapy for squamous cell carcinoma of the maxillary sinus. Jpn J Clin Oncol. 2005;35(11):633–638. Lee YY, Dimery IW, Van Tassel P, et al. Superselective intraarterial chemotherapy of advanced paranasal sinus tumors. Arch Otolaryngol Head Neck Surg. 1989;115(4):503–511. Papadimitrakopoulou VA, Hong WK. Combined-modality therapy in paranasal sinus cancer. Cancer J Sci Am. 1999;5(4):208–210. Madison ML II, Sorenson JM, Samant S, et al. The treatment of advanced sinonasal malignancies with pre-operative intra-arterial cisplatin and concurrent radiation. J Neurooncol. 2005;72(1):67– 75. Adelstein DJ. Induction chemotherapy in head and neck cancer. Hematol Oncol Clin North Am. 1999;13(4):689–698, v–vi. Cohen EE, Lingen MW, Vokes EE. The expanding role of systemic therapy in head and neck cancer. J Clin Oncol. 2004;22(9):1743– 1752. Jeremic B, Shibamoto Y, Milicic B, et al. Hyperfractionated radiation therapy with or without concurrent low-dose daily cisplatin in locally advanced squamous cell carcinoma of the head and neck: A prospective randomized trial. J Clin Oncol. 2000;18(7):1458–1464. Brizel DM, Albers ME, Fisher SR, et al. Hyperfractionated irradiation with or without concurrent chemotherapy for locally advanced head and neck cancer. N Engl J Med. 1998;338(25):1798– 1804. Wendt TG, Grabenbauer GG, Rodel CM, et al. Simultaneous radiochemotherapy versus radiotherapy alone in advanced head and neck cancer: A randomized multicenter study. J Clin Oncol. 1998;16(4):1318–1324. Huguenin P, Beer KT, Allal A, et al. Concomitant cisplatin significantly improves locoregional control in advanced head and neck cancers treated with hyperfractionated radiotherapy. J Clin Oncol. 2004;22(23):4665–4673. Pignon JP, Bourhis J, Domenge C, et al. Chemotherapy added to locoregional treatment for head and neck squamous-cell carcinoma: Three meta-analyses of updated individual data. MACHNC Collaborative Group. Meta-Analysis of Chemotherapy on Head and Neck Cancer. Lancet. 2000;355(9208):949–955. Bernier J, Domenge C, Ozsahin M, et al. Postoperative irradiation with or without concomitant chemotherapy for locally advanced head and neck cancer. N Engl J Med. 2004;350(19):1945–1952. Cooper JS, Pajak TF, Forastiere AA, et al: Postoperative concurrent radiotherapy and chemotherapy for high-risk squamous-cell carcinoma of the head and neck. N Engl J Med. 2004;350:1937– 1944. Bernier J, Cooper JS, Pajak TF, et al: Defining risk levels in locally advanced head and neck cancers: a comparative analysis of concurrent postoperative radition plus chemotherapy trials of the EORTC (#22931) and RTOG (#9501). Head Neck. 2005;27:843– 850. Monnerat C, Faivre S, Temam S, et al. End points for new agents in induction chemotherapy for locally advanced head and neck cancers. Ann Oncol. 2002;13(7):995–1006. Hitt R, Lopez-Pousa A, Martinez-Trufero J, et al. Phase III study comparing cisplatin plus fluorouracil to paclitaxel, cisplatin, and fluorouracil induction chemotherapy followed by chemoradiotherapy in locally advanced head and neck cancer. J Clin Oncol. 2005;23(34):8636–8645. Vermorken J, Remenar E, van Herpen C, et al. Standard cisplatin/infusional 5-fluorouracil (PF) vs docetaxel (T) plus PF (TPF) as neoadjuvant chemotherapy for nonresectable locally advanced squamous cell carcinoma of the head and neck: A phase III trial of the EORTC Head and Neck Cancer Group (EPORTC #24971). N Engl J Med. 2007;357(17):1695–1704. Posner MR. Sequential combined modality therapy a phase III study: TAX 324 TPF vs PF followed by chemoradiotherapy.
Chapter 11: Chemotherapy for Skull Base Tumors
28.
29.
30.
31.
32.
33. 34.
35.
36.
37.
38.
39.
40.
Special presentation. American Society of Clinical Oncology Meeting; Atlanta, GA, N Engl J Med. 2007;357(17):1705–1715. Levine PA, Frierson HF Jr, Stewart FM, et al. Sinonasal undifferentiated carcinoma: A distinctive and highly aggressive neoplasm. Laryngoscope. 1987;97(8)(Pt 1):905–908. Ejaz A, Wenig BM. Sinonasal undifferentiated carcinoma: Clinical and pathologic features and a discussion on classification, cellular differentiation, and differential diagnosis. Adv Anat Pathol. 2005;12(3):134–143. Rischin D, Porceddu S, Peters L, et al. Promising results with chemoradiation in patients with sinonasal undifferentiated carcinoma. Head Neck. 2004;26(5):435–441. Enepekides DJ. Sinonasal undifferentiated carcinoma: An update. Curr Opin Otolaryngol Head Neck Surg. 2005;13(4):222– 225. Svane-Knudsen V, Jorgensen KE, Hansen O, et al. Cancer of the nasal cavity and paranasal sinuses: A series of 115 patients. Rhinology. 1998;36(1):12–14. Resto VA, Eisele DW, Forastiere A, et al. Esthesioneuroblastoma: The Johns Hopkins experience. Head Neck. 2000;22(6):550–558. Polin RS, Sheehan JP, Chenelle AG, et al. The role of preoperative adjuvant treatment in the management of esthesioneuroblastoma: The University of Virginia experience. Neurosurgery. 1998;42(5):1029–1037. McLean JN, Nunley SR, Klass C, et al. Combined modality therapy of esthesioneuroblastoma. Otolaryngol Head Neck Surg. 2007;136(6):998–1002. Jeng YM, Lin CY, Hsu HC. Expression of the c-kit protein is associated with certain subtypes of salivary gland carcinoma. Cancer Lett. 2000;154(1):107–111. Freier K, Flechtenmacher C, Walch A, et al. Differential KIT expression in histological subtypes of adenoid cystic carcinoma (ACC) of the salivary gland. Oral Oncol. 2005;41(9):934–939. Mori Y, Kobayashi T, Kida Y, et al. Stereotactic radiosurgery as a salvage treatment for recurrent skull base adenoid cystic carcinoma. Stereotact Funct Neurosurg. 2005;83(5–6):202–207. Esmaeli B, Golio D, Kies M, et al. Surgical management of locally advanced adenoid cystic carcinoma of the lacrimal gland. Ophthal Plast Reconstr Surg. 2006;22(5):366–370. Licitra L, Grandi C, Prott FJ, et al. Major and minor salivary glands tumours. Crit Rev Oncol Hematol. 2003;45(2):215–225.
179
41. Haddad RI, Posner MR, Busse PM, et al. Chemoradiotherapy for adenoid cystic carcinoma: Preliminary results of an organ sparing approach. Am J Clin Oncol. 2006;29(2):153–157. 42. Samant S, van den Brekel M, Kies M, et al. Concurrent chemoradiation for adenoid cystic carcinoma of the head and neck: A 3-institution experience. Poster presented at the 7th International Conference on Head and Neck Cancer, July 19–23, 2008, San Francisco, CA. 43. Spiro RH. Distant metastasis in adenoid cystic carcinoma of salivary origin. Am J Surg. 1997;174(5):495–498. 44. Fordice J, Kershaw C, El-Naggar A, et al. Adenoid cystic carcinoma of the head and neck: Predictors of morbidity and mortality. Arch Otolaryngol Head Neck Surg. 1999;125(2):149–152. 45. Dreyfuss AI, Clark JR, Fallon BG, et al. Cyclophosphamide, doxorubicin, and cisplatin combination chemotherapy for advanced carcinomas of salivary gland origin. Cancer. 1987;60(12):2869– 2872. 46. Dodd RL, Slevin NJ. Salivary gland adenoid cystic carcinoma: A review of chemotherapy and molecular therapies. Oral Oncol. 2006;42(8):759–769. 47. Glisson B, Blumenschein G, Francisco M, et al. Phase II trial of gefitinib in patients with incurable salivary gland cancer. J Clin Oncol. 2005;23:508S; Abstr 5532. 48. Hotte SJ, Winquist EW, Lamont E, et al. Imatinib mesylate in patients with adenoid cystic cancers of the salivary glands expressing c-kit: A Princess Margaret Hospital phase II consortium study. J Clin Oncol. 2005;23(3):585–590. 49. Silva EG, Butler JJ, Mackay B, et al. Neuroblastomas and neuroendocrine carcinomas of the nasal cavity: A proposed new classification. Cancer. 1982;50(11):2388–2405. 50. Babin E, Rouleau V, Vedrine PO, et al. Small cell neuroendocrine carcinoma of the nasal cavity and paranasal sinuses. J Laryngol Otol. 2006;120(4):289–297. 51. Fitzek MM, Thornton AF, Varvares M, et al. Neuroendocrine tumors of the sinonasal tract. Results of a prospective study incorporating chemotherapy, surgery, and combined proton-photon radiotherapy. Cancer. 2002;94(10):2623–2634. 52. Bhattacharyya N, Thornton AF, Joseph MP, et al. Successful treatment of esthesioneuroblastoma and neuroendocrine carcinoma with combined chemotherapy and proton radiation. Results in 9 cases.Arch Otolaryngol Head Neck Surg. 1997;123(1):34–40.
12 Rehabilitation of Speech and Swallowing of Patients with Tumors of the Skull Base Gail L. Davie, Denise A. Barringer, and Jan S. Lewin
INTRODUCTION
NORMAL NEUROPHYSIOLOGY OF SWALLOWING AND COMMUNICATION
The inability to communicate or eat normally are two of the most disturbing outcomes of head and neck, brain, and skull base tumors (1). Both communication and swallowing are complex processes that require a highly specific interplay between the efferent signals from the cortex, subcortex, and brainstem and the afferent signals from the structures of the aerodigestive tract, including the pharynx, larynx, and oral cavity. Skull base tumors, regardless of the site of the lesion or its pathology, put patients at risk for the development of a variety of communication and/or swallowing deficits. Despite lower morbidity and mortality rates resulting from recent advances in the medical treatment of skull base tumors, the frequency and severity of residual deficits in both communication and alimentation have changed little (2). As a result, many of these patients, although cured of their disease, become social recluses because of their functional disabilities. Early preoperative consultation with speech-language pathologists who are expert in the management of patients with skull base tumors is critical. Data show that optimal recovery is facilitated when the patient and family receive important information regarding the disease and its effect on survival and posttreatment quality of life (3). Critical information provided by an experienced speech-language pathologist helps the medical team prepare the patient and family for the potential effects of treatment on later speech and swallowing function. In addition, experience has shown that patients who are knowledgeable are often better able to accept their treatment and the potential for functional consequences. Therefore, it is important that patients with skull base tumors and their families thoroughly understand the crucial need for posttreatment rehabilitation. This chapter highlights critical components in the evaluation and rehabilitation of patients with communication and swallowing disorders associated with tumors of the skull base. To enhance the reader’s understanding of this complex topic, we have divided the chapter into three areas of focus: (i) normal neurophysiology of swallowing and communication, (ii) common swallowing and communication disorders associated with tumors of the base of skull and their treatment, and (iii) assessment and treatment of communication and swallowing disorders in patients with tumors of the base of skull.
Swallowing Swallowing is a complex interaction of biomechanical, neurophysiologic, and behavioral events that occur in strict hierarchic sequence and depend upon intact sensory awareness and basic recognition of the act of eating. Historically, swallowing was considered solely a brainstem function of sensory and motor integration coordinated by the reticular nuclei within the pons and medulla (4–5). However, recent research using functional magnetic resonance imaging has shown broad cortical involvement during swallowing, with the most prominent centers of activity located within the lateral precentral and postcentral gyri and the right insula (6). Although the sequence of muscle activity in swallowing generally occurs once the pharyngeal swallow has been triggered; sensory feedback may alter the precise occurrences of muscle contractions. Sensory innervation to the mouth, pharynx, and larynx is generally provided by the trigeminal (cranial nerve V), glossopharyngeal (cranial nerve IX), and vagus (cranial nerve X) nerves. Each of these nerves provides important sensory information that helps initiate and control the act of swallowing. Motor innervation in swallowing is mediated primarily via cranial nerves V, IX, X, and XII (hypoglossal) (7). Many of the techniques used for swallowing rehabilitation rely on this sensorimotor interrelationship, utilizing sensory pathways to stimulate motor function (8). Swallowing is generally described in four phases or stages: oral preparatory, oral, pharyngeal, and esophageal. The sequence of these stages is generally predictable, although the relative timing varies depending on the size and texture of the item being swallowed. It is the responsibility of the speech pathologist to evaluate the first three phases, while the evaluation of esophageal anatomy and disorders is the responsibility of the gastrointestinal radiologist. Thus, esophageal functioning is not discussed within this chapter and the reader is referred elsewhere for information. Swallowing has been recognized as a largely reflexive activity, as only the oral preparatory and oral phases are purely voluntary. The oral preparatory phase initiates the process of swallowing. During this phase, smooth coordination and transition between mastication and manipulation prepare and form the food into a cohesive, manageable bolus. The oral stage of swallowing begins as the tongue starts to
181
182
Davie et al.
move in an anterior to posterior rolling or stripping motion, propelling the bolus posteriorly to the oropharynx. The oral stage of swallowing typically lasts less than 1.5 seconds, but may take slightly longer with increased bolus viscosity. The pharyngeal phase of swallowing occurs with the initiation or “trigger” of the swallowing reflex. Generally, the area between the anterior faucial arches and the point at which the tongue base crosses the rim of the mandible is considered to be the key location for initiation of the pharyngeal swallow. The swallow trigger primarily depends on cranial nerve IX. In synchrony with the trigger of the swallow reflex, the soft palate elevates and retracts to close the velopharyngeal port to prevent nasal regurgitation of swallowed material. The base of tongue makes contact with the bulging posterior pharyngeal wall to help propel the food through the pharynx, while the larynx and hyoid bone move superiorly and anteriorly to prevent food from entering the airway, thereby directing the bolus posteriorly into the cervical esophagus. The elevation and anterior movement of the hyolaryngeal complex are essential to the opening of the upper esophageal sphincter or cricopharyngeus muscle, which allows the bolus to enter the cervical esophagus without aspiration. The pharyngeal phase also typically lasts less than 1.5 seconds, but unlike the oral preparatory and oral stages, the pharyngeal phase retains both voluntary and involuntary mechanisms of control.
Communication Analogous to swallowing, the complexity of communication is often not appreciated until it is lost. Communication is more than simply the articulation of sounds and syllables into meaningful utterances, or speech. Communication begins within the cerebral cortex and subcortex as a thought, an internal representation of an experience. Within the perisylvian region, specifically Broca area and the insula, the thought is organized and converted into language, a rulegoverned system of symbols represented by words that are eventually converted into speech. The sounds of the language are applied and the motor program, or directions for speech production, is generated. The motor program or plan is trans-
ferred from the premotor cortex to the lower portion of the motor cortex through tracts to the cranial nerves for activation of the muscles of the respiratory tract, larynx, and oral cavity into speech (9). It is important to remember that the network of association tracts communicates between critical areas of the brain and is essential to maintaining the speed and accuracy of normal communication. Therefore, any disruption or injury, focally or indirectly, to these regions or tracts will result in communication problems. The type of communication problem depends on the particular process or processes affected.
COMMON DISORDERS OF SWALLOWING AND COMMUNICATION Disorders of Swallowing Swallowing disorders associated with tumors of the skull base are not generally the result of a single cause; rather, they are the culmination of multiple insults resulting from the tumor itself, the surgical resection, the surgical approach, and/or the adjuvant treatments used to cure the disease. In general, tumors that affect the posterior skull base will result in more profound swallowing dysfunction because the cranial nerves critical to swallowing originate in this region. Table 1 summarizes the swallowing deficits and salient features associated with each cranial nerve. The consequences of damage to these nerves are often devastating, and many patients require long-term use of a gastrostomy tube for nutrition and a permanent tracheotomy tube because of chronic aspiration, which can lead to aspiration pneumonia, a frequent and serious problem in patients with skull base tumors. In addition to the posterior skull base, the anterior and middle regions of the skull base each have specific tumor- and treatment-associated swallowing disorders.
Anterior Skull Base Swallowing disorders associated with anterior skull base tumors occur as a result of anatomic alterations rather than cranial nerve dysfunction. The surgical exposure of anterior skull base tumors may require resection of structures such as the
Table 1 Speech and Swallowing Impairments Associated with Each Cranial Nerve Cranial nerve CN V CN VII CN VIII
CN IX
CN X
CN XII
Disordered physiology
Swallowing symptoms
• ↓ jaw movement • ↓ face, mouth and jaw sensation • ↓ facial movement and sensation • ↑ salivation • ↓ hearing • ↓ balance
• ↓ mastication • ↓ oral containment • Residue in the lateral sulci • Drooling N/A
• Delayed pharyngeal trigger • ↓ velopharyngeal closure • ↓ laryngeal elevation • ↓ palatal, pharyngeal, laryngeal excursion
• Aspiration before and during the swallow • Stasis/residue in the valleculae, posterior pharyngeal wall and pyriform sinuses • Aspiration during or after the swallow
• ↓ true vocal cord abduction and/or adduction • ↓ pharyngeal sensation
• Inability to cough
• ↓ lingual range of motion and strength
• ↓ bolus consolidation • ↓ anterior to posterior movement of the bolus • Oral residue
Speech symptoms • Unilateral: insignificant • Bilateral: severely ↓ articulatory precision • Mild distortion of b, p, f, v, • Distortion of resonance • ↓ articulatory precision of all sounds over time • Hypernasality • Hypernasality • Breathiness and hoarseness • ↓ pitch range • ↓ vocal loudness • Imprecise articulation of l, t, d, s, z, sh, ch, k, g
Chapter 12: Rehabilitation of Speech and Swallowing of Patients with Tumors of the Skull Base
maxilla or mandible that provide important structural support during swallowing. Therefore, functional deficits most commonly occur in the oral preparatory and/or oral phases of swallowing. In general, anterior skull base tumors do not affect the pharyngeal stage of swallowing. Labial sensory loss resulting from splitting of the maxilla often renders patients unable to maintain a labial seal, which results in drooling and food loss from the mouth. More extensive surgical approaches such as total maxillectomy cause palatal defects that hinder oral transit, lingual movements and contacts, and often result in velopharyngeal incompetency and nasal regurgitation. These deficits can be extremely detrimental to patient’s quality of life, preventing them from eating in public, and in the extreme, from eating by mouth.
Middle Skull Base Like tumors of the anterior skull base, tumors of the middle skull base are more apt to affect the oral preparatory and oral phases of swallowing than the pharyngeal phase. Damage to the middle skull base can result in unilateral or bilateral impairment of cranial nerves V and/or VII, which control the ability to open the mouth, masticate, and to maintain facial tone and symmetry. Generally, patients are able to compensate for unilateral damage affecting one or both nerves. However, bilateral damage can lead to profound dysfunction in the oral preparatory phase owing to the inability to close the mouth, maintain buccal tension, or engage in rotary chewing.
Posterior Skull Base Although anterior and middle skull base tumors can cause significant swallowing difficulties and impede the speed, adequacy, and efficiency of oral intake, patients can usually compensate for these deficits. However, tumors that affect the posterior skull base generally result in significantly more severe deficits and present greater challenges for rehabilitation of swallowing function. In most cases, injuries to cranial nerves IX and X occur together because both nerves exit the base of skull through the jugular foramen. Injury to cranial nerve IX impairs the pharyngeal swallowing reflex, pharyngeal contraction, and velopharyngeal competency. The pharyngeal swallowing reflex may be delayed or even absent, resulting in aspiration before or during the swallow. Neural injury will impair normal pharyngeal contraction and velopharyngeal competency that will result in pharyngeal stasis of food after the swallow and nasal regurgitation. Injury to the vagus nerve is often even more debilitating, as this nerve is responsible for several physiologic activities that work in concert with the pharyngeal trigger of swallowing. The rate of postoperative vagal nerve injuries has been reported to be as high as 50% after specific approaches (10). High vagal injuries result in the most serious problems because they affect all three branches of the nerve: the pharyngeal, superior laryngeal, and recurrent laryngeal nerves (11). Therefore, patients with high vagal injuries will have difficulty moving food through the pharynx and aspirate as a result of injury to all three branches. Impairment at any level of the nerve may result in aspiration because of decreased laryngeal sensation, impaired pharyngeal contraction, or reduced airway closure. Clinically, the superior and recurrent laryngeal branches of the vagus nerve are most important for glottic airway protection, while the pharyngeal branch ensures adequate pharyngeal contraction via innervation to the pharyngeal constrictor musculature. Insults to the superior laryngeal nerve can leave patients insensate to aspirate, caus-
183
ing them to silently, without coughing or any other indication of awareness, aspirate food into the trachea. Damage to the recurrent laryngeal nerve usually results in paresis or paralysis of one or both of the true vocal cords, depending upon the level of injury, impairing glottic valving in response to the aspirate. Clinicians should be advised that patients with vagal injuries are at high risk for aspiration pneumonia, and a lack of patient reaction should not be equated with safe swallowing in this population at high risk for silent aspiration. The tongue is one of the most critical structures in the oropharyngeal swallow. Any damage, unilateral or bilateral, to cranial nerve XII, the hypoglossal nerve, affects the lingual movements critical to the manipulation and transport of food from the anterior portion of the oral cavity to the pharynx. Patients with hypoglossal nerve injuries will have difficulty maintaining a cohesive food bolus during the oral stage of swallowing for efficient transport to the oropharynx. More importantly, base of tongue retraction to the posterior pharyngeal wall propels the food through the pharynx. If this movement is impaired, a significant amount of food collects within the valleculae, resulting in swallowing that is laborious, lengthy, and inefficient. Patients may thus expend an inordinate amount of energy for little nutritional gain. In addition, these patients remain at risk for aspiration because of the food that is left in the pharynx after the swallow. Cranial nerve injury is the most common cause of dysphagia associated with skull base lesions, with reports of aspiration in as many as 75% of patients (12). The cranial nerves affected and the level of injury determine the type and severity of the swallowing disorder. Furthermore, patients with gradual tumor progression generally will demonstrate the best potential for compensation of swallowing deficits, whereas patients who experience acute insults will have more severe problems because they have not had the benefit of time to adjust to their swallowing problems (2). It is therefore critical that patients with cranial nerve dysfunction be referred to a speech-language pathologist for early identification of dysphagia to maintain oral nutrition for as long as possible and to avoid further complications in this high-risk population.
Disorders of Communication The most common disorders of communication associated with skull base lesions and their treatment are those that affect the motor aspects of speech production rather than meaning or language. Dysarthria remains the primary communication impairment in this patient population. It refers to a group of speech disorders characterized by weakness, paralysis, incoordination, sensory deprivation, and alterations in muscle tone (13). Dysarthric speech is often described as slurred and imprecise. There are six subtypes of dysarthria, each of which is distinguished on the basis of its auditory perceptual characteristics, which in turn have diagnostic implications for lesion localization (14). The ability to differentiate between these subtypes requires specialized training and experience. Therefore, it is critical that patients be referred to speech pathologists who are expert in the differential diagnosis of dysarthria to ensure accuracy in both diagnosis and treatment.
Anterior Skull Base As in swallowing, communication impairments associated with the anterior skull base are generally a function of damage to the anatomic structures. For example, maxillectomy results in a palatal defect that impairs vocal resonance, or the quality of speech, as well as the overall intelligibility of
184
Davie et al.
speech. Palatal prostheses or obturators that fill the defect and reestablish anatomic continuity are generally the best way to manage these types of problems. When mandibular defects are involved, patients experience difficulty opening and closing their mouths and producing such sounds as /b/, /p/, and /m/. It is important to note that speech impairments that result from structural damage are not considered dysarthric, as they are not neurogenic in origin. Less commonly, larger lesions of the anterior skull base that require retraction of the frontal lobes may also result in language disorders. The most common of these are aphasia and right hemisphere disorder (RHD). Aphasia that affects verbal expression, writing, auditory comprehension, and reading is generally an outcome of insult to the perisylvian cortex of the dominant hemisphere or subcortical damage to the thalamus and basal ganglia (15). In contrast, RHD is a disorder of pragmatics or the social use of language. Unlike patients with aphasia or dysarthria, patients who have RHD exhibit appropriate use of grammar and words without misarticulations of speech, but their communication is nonspecific and inefficient. These patients frequently say things that are irrelevant, disconnected, incompatible with the situation, and produced without appropriate affect (16). The impact of such deficits may appear subtle but can be devastating to interpersonal relationships.
Middle/Posterior Skull Base Communication disorders that affect both the middle and the posterior base of skull result from cranial nerve dysfunction. Primary among such disorders is flaccid dysarthria, a subtype of dysarthria characterized by weakness of the muscles responsible for speech production, occurring as a result of cranial nerve dysfunction (14). Flaccid dysarthria occurs with similar frequency in patients with middle and posterior skull base tumors, with 14% of all reported cases, including those occurring from stroke, head injury, etc., attributed to surgical procedures involving the middle and posterior skull base regions (14). When lesions involve the cerebellum, the dysarthria will present as ataxic. In contrast to flaccid dysarthria, ataxic dysarthria is characterized primarily by irregularity of articulatory movements. Ataxic dysarthria resulting from brainstem tumors accounts for 3% of all reported cases (14). Damage to any of the lower cranial nerves will result in flaccid dysarthria. The resulting speech impairments will generally be consistent; however, the most prominent feature will depend upon the specific cranial nerve(s) involved and the extent of damage, unilateral or bilateral. Even skull base surgeries that do not result in nerve resection can damage nerves. Prolonged retraction or stretching of the nerve may result in deficits that cannot be restored completely (2). The critical nerves in the lower region for speech are cranial nerves V, VII, IX, X, and XII. Another lower cranial nerve, cranial nerve VIII, is not generally considered a nerve for speech but is imperative for hearing. Therefore, damage to this nerve, which is most often a result of acoustic neuromas and other types of tumors that affect the vestibular nerve, can be traumatic to both speech and communication. This topic, however, is beyond the scope of this chapter, and the reader is referred to the literature regarding vestibular tumors and hearing for a comprehensive review. The close proximity of cranial nerves IX and X makes it difficult to discern the impact of damage to cranial nerve IX on speech production, as it is rarely damaged in isolation
(11). Likely, the effect is on both resonance and phonation, but the importance of cranial nerve IX in the assessment of dysarthria is indeterminate for practical purposes (14). The effects of damage to cranial nerve X are more complicated because of its long course and its three major branches. Hypernasality, audible nasal emission, and imprecise production of consonants that require high oral pressure are usually the result of velopharyngeal incompetence owing to damage to the pharyngeal branch of the vagus nerve. Unilateral lesions that include the superior and recurrent laryngeal branches often result in breathiness or aphonia, reduced vocal intensity, and limited pitch range. Lesions of the superior laryngeal nerve that spare the pharyngeal and recurrent branches are most often associated with an inability to change the pitch of the voice. Sometimes mild breathiness or hoarseness may also be present. Unilateral recurrent laryngeal nerve lesions resulting in impaired true vocal fold motion cause weak vocal intensity, hoarse vocal quality, problems with pitch, and sometimes diplophonia, or double sound. Bilateral damage may cause inhalatory stridor or airway obstruction, depending on the type of vocal fold paralysis, abductor or adductor. Damage to cranial nerve XII can significantly affect speech production. The overriding speech characteristic in unilateral and bilateral XII nerve lesions is imprecise articulation, specifically of sounds involving the tongue. With unilateral lesions, compensation is generally adequate to allow for speech perception. Bilateral lingual weaknesses, however, result in greater articulatory distortion because of the inability to achieve the rapid movements and precise articulatory placements required for intelligible speech production. Direct or indirect damage to the cerebellum or surrounding regions, including the cerebellar pontine angle and the dentate nucleus, frequently result in ataxic dysarthria. In contrast to flaccid dysarthria, ataxic dysarthria’s hallmark feature is irregularity. Speech characteristics can include altered prosody, prolongation of sounds, and intermittent disintegration of articulation with irregular changes in pitch and loudness. This is most commonly described as “drunken” speech (14). One of the more unusual communication disorders associated with late presentation of severe ataxic dysarthria, cerebellar mutism syndrome (CMS), occurs mainly in children, without associated cranial nerve damage or long tract signs. CMS is characterized by a transient mutism that occurs within 48 hours after resection of the cerebellar mass lesion and may last up to 6 months, followed by severe dysarthria (17,18). Speech pathology services have been shown to facilitate recovery of function in patients whose symptoms persist beyond 8 weeks (19). CMS should also be considered in adults with large cerebellar lesions or with medulloblastomas greater than 5 cm in diameter (20,21).
ASSESSMENT AND INTERVENTION FOR SWALLOWING AND COMMUNICATION DISORDERS Evaluation of swallowing and communication should be performed as early as possible in patients with pretreatment tumor-related deficits and, preferably, before treatment begins in patients who are at high risk for treatment-related dysfunction. A comprehensive evaluation should include a thorough review of the patient’s history, a patient/family interview, a neurologic screening that includes an examination of the oral mechanism, and a speech profile. Formal measures
Chapter 12: Rehabilitation of Speech and Swallowing of Patients with Tumors of the Skull Base
185
provide additional information for differential diagnosis and lesion localization. The results from these examinations are carefully analyzed, and an individually tailored treatment plan specific to the presenting communication and swallowing symptoms is designed.
Assessment Swallowing Clinical Assessment The clinical or bedside swallowing examination allows observation of important events in the oral stage of swallowing, including labial and lingual control, palatal function, and oral sensitivity. It affords an opportunity to observe the patient’s reactions to swallowing and to assess the impact of behaviors such as attention, impulsivity, and judgment on the ability to eat safely. Clinical swallowing examinations provide important information regarding the need for further objective swallowing assessments such as the modified barium swallow study or the fiberoptic endoscopic evaluation of swallowing. Clinical assessments of swallowing are not reliable indicators of pharyngeal physiology, and therefore, should not be used as a replacement for objective swallowing measures. Instrumental Assessment There are two main instrumental studies of swallowing, the modified barium swallow (MBS) study and the fiberoptic endoscopic evaluation of swallowing (FEES). The indications for use of each test depend upon the presenting swallowing problem and the patient’s ability to undergo a radiographic examination. Modified Barium Swallow Study Perhaps the most widely used tool to evaluate oropharyngeal swallowing is the videofluoroscopic examination, or MBS study (22,23). The MBS study is different from the standard barium swallow in that it shows the entire oropharyngeal swallow, including all four stages of swallowing. It provides information regarding all events, allowing for the diagnosis of specific swallowing disorders and causes of aspiration. A further benefit is that it gives the clinician the opportunity to attempt specific therapeutic techniques during the study to determine their ability to alleviate the swallowing deficit.
Figure 1 Aspiration on an MBS study.
agnosing speech disorders such as dysarthria. Perceptual speech characteristics provide powerful information for accurately differentiating between speech disorders. When perceptual analyses are administered and interpreted properly, they may also provide critical information to help identify specific neurologic disease processes (14). It is important to remember, however, that accurate perceptual analysis is a difficult skill to master and requires significant training, experience, and practice that few clinicians achieve. In contrast, disorders of language such as aphasia and RHD are best evaluated using standardized test measures such as the Western Aphasia Battery and the Mini Inventory of Right Brain Injury-2 (25,26). These tests rely less on perceptual, subjective judgments and more on the objective analysis and documentation of the semantic and syntactic errors that commonly occur in patients with language disorders.
Treatment In general, treatment for speech and swallowing disorders either restores function or compensates for dysfunction. Therapeutic exercises that improve the strength and coordination of the oropharyngeal musculature are important
Fiberoptic Endoscopic Evaluation of Swallowing FEES is a videoendoscopic tool that can be used at the bedside or in the clinic by a single clinician (24). It is the key assessment tool for visualization of aspiration, airway protection, and laryngeal sensation. There are two distinct disadvantages associated with FEES, however. First, visualization of the oral stage of swallowing is obstructed during FEES. Second, much of the oropharyngeal physiology of swallowing must be inferred because of the limitations of endoscopy. Figures 1 and 2 compare aspiration as evaluated via MBS and FEES.
Communication The examination of communication includes a combination of formal standardized test batteries and informal perceptual measures that are selected on the basis of the patient’s presenting problems. Informal test measures that evaluate the rate, rhythm, and precision of speech production may include the repetition of sounds in isolation, the production of rapidly sequenced syllables, and maximum phonation duration. These measures, along with examination of the oral mechanism, are the most useful tools for differentially di-
Figure 2 Aspiration on an FEES examination.
186
Davie et al.
restorative techniques for both speech and swallowing disorders. Regardless of the potential for functional improvement, compensatory strategies are always important in the treatment of speech and swallowing impairments because they provide the patient with alternative ways to maintain their normal routine in the short term while working on long-term functional improvement. Compensatory strategies become the focus of treatment when full restoration of normal function is not a realistic goal. In cases of swallowing impairment, compensatory approaches control the flow of food and eliminate the patient’s symptoms such as aspiration, but do not necessarily change the physiology of the patient’s swallow (8). Examples of such approaches include postural changes, diet modifications, special feeding techniques, and the use of intraoral prosthetic devices. Patients with communication impairments, for example dysarthria, benefit from compensatory strategies such as a slow rate of speech or the use of an alternative communication device, a picture board or a voice-generated system, to improve listener understanding and reduce frustration. Aphasic patients may employ gestures or simplify their message to compensate for their language deficits.
CASE STUDIES Case Study 1 The patient was a 55-year-old right-handed woman with a diagnosis (in 1972) of anaplastic ependymoma of the fourth ventricle. She was status post three subtotal resections as well as radiation treatment to the posterior fossa for a total of 54 Gy (in 1991). The patient was eating a regular diet including both solids and liquids. She subsequently underwent percutaneous fluoroscopic gastrostomy in November of 1995 due to persistent dysphagia and recurrent episodes of pneumonia. She presented to the speech pathology service one year later (1996) with complaints of a progressive decline in speech and swallowing. An MBS study and speech assessment were performed. Clinical assessment revealed multiple cranial nerve palsies (III, V, VII, IX, X, and XII) characterized by articulatory imprecision due to lingual weakness, hypernasality, irregular prosody, and dysphonia. Conversational speech was approximately 50% intelligible. Results revealed a moderate to severe mixed flaccid-ataxic dysarthria. The patient was taught to use compensatory speech strategies to improve intelligibility to communicate needs with family members. The results from the MBS study showed impaired lingual strength and coordination, decreased bolus formation, delayed initiation of the swallow reflex, reduced base of tongue retraction, and reduced hyolaryngeal excursion. Silent aspiration occurred before, during, and after the swallows of thin liquids. Compensatory strategies and diet modifications were unsuccessful in alleviating aspiration. Objective swallowing examination facilitated diagnosis of a moderately severe oropharyngeal dysphagia. Recommendations included omission of oral intake of liquids. Swallowing was judged adequate for other food consistencies. No further episodes of aspiration pneumonia were documented.
Case Study 2 The patient was a 59-year-old right-handed woman diagnosed with a giant left petroclival meningioma. There was no documentation of any cranial nerve impairments, dysarthria, or dysphagia at baseline reported by the neurology service.
Resection of the tumor was performed via a two-staged, left suboccipital craniotomy with placement of a right frontal ventriculostomy. A speech pathology consultation was generated postoperatively. Results revealed multiple cranial nerve impairments (IX, X, and XII). A clinical swallow evaluation revealed continual throat clearing and the use of multiple swallows while eating suggesting probable pharyngeal dysphagia and aspiration. Objective examination of swallowing via an MBS study was recommended. Findings showed delayed initiation of the swallow and an inability to propel the bolus through the pharynx because of weak pharyngeal motility. A percutaneous endoscopic gastrostomy was performed due to inability to maintain oral nutrition and hydration. The results of speech and swallowing examination revealed moderate-severe flaccid dysarthria with severe pharyngeal dysphagia. A speech and swallowing treatment plan was designed based on the findings from clinical and objective examinations and was implemented during the patient’s hospitalization. Patient and family education was provided regarding treatment objectives to ensure realistic expectations for recovery and compliance with the treatment plan. The patient continued to receive speech and swallowing services following discharge. The patient presented for reevaluation three months postdischarge. Findings showed no significant swallowing deficits. The patient was able to eat by mouth and the percutaneous endoscopic gastrostomy tube was removed. Although dysarthria improved slightly, dysphonia persisted as a result of cranial nerve X dysfunction.
CONCLUSION Patients with skull base tumors present significant diagnostic, therapeutic, and rehabilitation challenges. Because the treatment is so complex and the potential for complications so high, a strong multidisciplinary team made up of specialists from many fields who work together is needed for the patient to benefit from the unique expertise of each team member. The diagnosis of a brain tumor creates turmoil and fear in both patients and their families. In addition, the acute effects of a skull base tumor can prevent normal communication, impede the ability to eat by mouth, and severely disrupt normal patient and family interactions, the effects of which are generally overwhelming. Whether the deficits are acute or long-term, the speech pathologist plays a critical role in providing information about normal function and rehabilitation, identifying and teaching compensatory strategies to the patient and family to preserve normal communication as much as possible, and providing appropriate intervention to help maintain oral nutrition while avoiding medical complications such as aspiration. It is, therefore, important that the speech pathologist be consulted early in the patient’s course of evaluation and treatment to give the patient and the family the opportunity to ask questions and to better understand information that at first seems overpowering and frightening, thus reducing the fears and misconceptions associated with skull base tumors. Appropriate and timely referral to a speech pathologist with expertise in the rehabilitation of skull base tumor patients helps patients and their families maintain realistic expectations about their functional recovery and provide a rehabilitative plan so that optimal communication and swallowing outcomes can be achieved and the best quality of life attained.
Chapter 12: Rehabilitation of Speech and Swallowing of Patients with Tumors of the Skull Base REFERENCES 1. Yorkston KM, Miler RM, Stand EA. Management of Speech and Swallowing in Degenerative Diseases. Austin, Texas: Pro-Ed, Inc, 1995. 2. Levine TM. Swallowing disorders following skull base surgery. Otolaryngol Clin North Am. 1988;21:751–759. 3. Portnoy RK. Head and Neck Cancer. New York, NY: Plenum Press, 1995:218–231. 4. Carptenter DO. Central nervous system mechanisms in deglutition and emesis. In: Schultz SG, ed. Handbook of Physiology. Gastrointestinal System Control of Food and Water Intake. Sect 6, Vol 1. Bethesda, MD: American Physiological Society, 1989:685– 714. 5. Jean A. Brainstem control of swallowing: Localization and organization of the central pattern generator for swallowing. In: Taylor A, ed. Neurophysiology of the Jaws and Teeth. New York, NY: MacMillan, 1990:294–321. 6. Martin RE, Goodyear BG, Gati JS, et al. Cerebral cortical representation of automatic and volitional swallowing in humans. J Neurophysiol. 2001;85:938–950. 7. Bhatnagar SC. Neuroscience for the Study of Communicative Disorders. Baltimore, Maryland: Lippincott Williams & Wilkins, 2002:333. 8. Logemann JA. Evaluation and Treatment of Swallowing Disorders. 2nd ed. Austin, TX: Pro-Ed, Inc, 1998. 9. Love RJ, Webb WG. Neurology for the Speech-Language Pathologist, 3rd ed. Boston, MA: Butterworth-Heinemann, 1996. 10. Cece JA, Lawson W, Biller HF, et al. Compilation in the management of large glomus jugulare tumors. Laryngoscope. 1987;97:152–157. 11. Eibling DE, Boyd EM. Rehabilitation of lower cranial nerve deficits. Otolaryngol Clin North Am. 1997;30:865–875. 12. Jennings KS, Siroky D, Jackson CG. Swallowing problems after excision of tumors of the skull base: Diagnosis and management in 12 patients. Dysphagia. 1992;7:40–44. 13. Dworkin JP. Motor Speech Disorders: A Treatment Guide. St. Louis, Missouri: Mosby, 1991.
187
14. Duffy JR. Motor Speech Disorders: Substrates, Differential Diagnosis, and Management. St. Louis, Missouri: Mosby-Year Book, Inc, 1995. 15. Rosenbek JC, LaPointe LL, Wertz RT. Aphasia: A Clinical Approach. Boston, Massachusetts: College-Hill Press, 1989:215. 16. Myers P. Right Hemisphere Damage: Disorders of Communication and Cognition. San Diego, California: Singular Publishing Group, Inc, 1999:4–5. 17. Van Dongen HR, Catsman-Berrevoets CE, Van Mourik M. The syndrome of “cerebellar” mutism and subsequent dysarthria. Neurology. 1994;44:2040–2046. 18. Pollack IF, Polinko P, Albright AL, et al. Mutism and pseudobulbar symptoms after resection of posterior fossa tumors in children: Incidence and pathophysiology. Neurosurgery. 1995;37:885–893. 19. Sherman JH, Sheehan JP, Elias JW, et al. Cerebellar mutism in adults after posterior fossa surgery: A report of 2 cases. Surg Neurol. 2005;63:476–479. 20. Catsman-Berrevoets CE, Van Dongen HR, Mulder PG, et al. Tumor type and size are high risk factors for the syndrome of “cerebellar” mutism and subsequent dysarthria. J Neurol Neurosurg Psychiatry. 1999;67:755–757. 21. Idlan F, Tuna M, Erman T, et al. The evaluation and comparison of cerebellar mutism in children and adults after posterior fossa surgery: Report of two adult cases and review of the literature. Acta Neurochir (Wien). 2002;144:463–473. 22. Logemann JA. Imaging the oropharyngeal swallow. Adm Radiology. 1993;3:20–24, 43. 23. Logemann JA. Normal Radiographic Anatomy and Physiology of the Oropharynx. In: Hyams H, ed. Manual for the Videofluoroscopic Study of Swallowing, 2nd ed.: Austin, TX: Pro-Ed, Inc, 1993. 24. Langmore SE, Schatz K, Alson N. Endoscopic and videofluoroscopic evaluations of swallowing and aspiration. Ann Otol Rhinol Larygol. 1991;100:678–681. 25. Kertz A. The Western Aphasia Battery. San Antonio, TX: The Psychological Corporation Harcourt Brace Jovanovich, Inc, 1982. 26. Pimental PA, Knight JA. Mini Inventory of Right Brain Injury (MIRBI-2). Austin, Texas: Pro-Ed, Inc, 2006.
13 Quality of Life of Patients with Skull Base Tumors Ziv Gil and Dan M. Fliss
ics related to the QOL of patients with skull base neoplasms: (1) the available instruments for estimating their QOL, (2) the impact of surgery on different aspects of patients’ QOL following skull base procedures, and (3) the means to improve the QOL of this patient population.
INTRODUCTION Quality of life (QOL) is usually defined as a construct that reflects the individual perception of overall well-being (1). There is more and more recognition of the importance of topics addressing QOL issues of patients with neoplasms. QOL is assessed in an effort to improve treatment modalities, to promote restoration of patient’s daily functioning, and to accelerate his/her return to normal life. Estimation of the influence of surgical procedures on QOL can serve as a mean by which the most appropriate surgical approach can be selected for a given patient. Detailed understanding of the different aspects of QOL may help surgeons improve assessment and management of patients, identify specific impediments as early as possible during follow-up, and apply specific medical interventions to patients with increased risk and poor outcome (1). Furthermore, early access of patients to detailed information about their disease can lead to better adjustment to an imminent medical condition. A multidimensional evaluation of QOL involves retrieving information on the physical, emotional, social, and economical aspects of the patient’s lifestyle, as well as on specific symptoms associated with the patient’s disease. Valid interpretation of the data requires generic questionnaires or disease-specific instruments, which cover the morbidity associated with the site of neoplasia and its treatment (2). The technical development of skull base surgery has had a major positive impact on the long-term survival of patients with lesions involving the anterior or lateral cranial fossa, the paranasal sinuses, and the orbits (3). Various techniques have been developed for resection of these tumors, among them are the craniofacial and subcranial approaches (frequently used for anterior skull base tumors) and the middle fossa, translabyrinthine, and infratemporal fossa approaches (for lateral skull base tumors)(4,5). These procedures may, however, carry a considerable risk as well as serious morbidity (6). Although surgical procedures and radiation therapy are crucial for the treatment of skull base neoplasms, tumor control should not be the only goal of patient care. During the last decade, numerous studies have assessed QOL issues in patients with head and neck or brain tumors (7,8) These questionnaires were directed toward cancer of the upper aerodigestive tracts or brain parenchyma, and they included items that were unrelated to anterior or lateral skull base tumor morbidity. These instruments also lacked questions on issues which are frequently encountered among patients undergoing anterior or lateral skull base procedures. An instrument which focuses on a specific disease has the advantage of increasing the chances that the information provided by the patients is uniform, and it improves responsiveness by including questions which are relevant only to the specific population being studied. This chapter focuses on three top-
Methods for Estimating Quality of Life Generic Instruments Few psychometrically valid generic instruments have been developed during the last decade for estimating the QOL of patients in general and that of oncological patients in specific. Table 1 shows a few examples of commonly used instruments for assessing QOL in the general population or in patients with neoplasia. One of the most popular among them is a short form containing 36 items, also known as the short form 36 (SF-36) questionnaire. This generic instrument was first published in 1992 and further validated worldwide and in different languages. It can be completed with the aid of an interviewer or as a self-administered questionnaire (8). The items included in the SF-36 questionnaire are divided into eight domains with overall physical and mental health component scores. These domains are physical functioning, role limitations physical, bodily pain, social functioning, general mental health, role limitations emotional, vitality and general health. One advantage of the SF-36 questionnaire is the existence of normative data and the form’s documented reliability and validity in different countries. It was used extensively for estimating the QOL of patients with acoustic neuromas (9–13). A shorter generic form is the SF-12, developed as a condensed version of the SF-36 questionnaire (14). SF-12 contains 12 items and takes half the time to complete compared to the SF-36 questionnaire (i.e., 5 and 10 minutes, respectively). Another generic instrument commonly used for estimating QOL in cancer patients is the European Organization for Research and Treatment of Cancer (EORTC) QOLQ-C30 (15,16). Similar to the SF-36, different versions of this questionnaire are available in other languages. Table 1 summarizes the advantages and disadvantages of some of the generic QOL instruments in general use.
Disease-Specific Instruments Any instrument for estimating QOL needs to cover the morbidity associated with the specific site of the tumor and to allow an accurate and sensitive description of the patient’s condition. During the last decade, numerous studies have assessed QOL issues in patients with head and neck or brain tumors (17–19). These questionnaires were directed toward cancer of the upper aerodigestive tracts and included items that were unrelated to anterior skull base tumor morbidity. Site-specific modules incorporating questions addressing disease- and treatment-specific problems have since been 189
190
Gil and Fliss
Table 1 Generic Instruments Name
Description
SF-36
The most commonly used instruments for estimation of QOL. Contain 36 items divided into 8 domains. Can be completed with the aid of an interviewer or as a self-administered questionnaire. A short version of the SF-36 questionnaire A self-administered instrument with 30 items. Different versions of this questionnaire are available in other languages. A generic self-administered instrument, developed for cancer patients. Contains 27 items. Multilingual translations available. One of the most widely used measures to assess quality of life. Contains 136 items and 12 subscales. Can be completed with the aid of an interviewer or as a self-administered questionnaire.
SF-12 EORTC QOLQ-C30
FACT-G
SIP30
Advantage
Disadvantage
Tested for reliability and validity. Normative data available.
Not specific for tumor patients. Lack disease-specific items.
Valid, rapid application. Tested for validity and reliability. A head and neck module available. Tested for reliability and validity, fast application.
Same as SF-36. Not validated for SB patients. Takes 30 min to complete.
Tested for reliability and validity.
Not tumor specific. Takes >30 min to complete.
Limited experience in clinical trials
Abbreviations: SF-36, short form 36; EORTC, QOLQ-C30, European organization for research and treatment of cancer and quality-of-life questionnaire; FACT-G, functional assessment of cancer therapy, general module; SIP30, sickness impact profile to assess quality of life; SB, skull base.
developed based on the EORTC QOLQ-C30 questionnaire for the head and neck. The head and neck module of the EORTC, however, includes questions on swallowing, breathing, and mastication, which are not relevant to symptoms of patients with skull base neoplasms. The absence of questions related to the specific symptoms of patients with skull base tumors limits the application of these instruments to this subpopulation of oncological patients. These instruments also lack questions on issues which are frequently encountered in patients undergoing skull base surgery. Table 2 summarizes the advantages and disadvantages of few of the available disease-specific instruments for QOL evaluation of patients with head and neck cancer. It should be borne in mind that patients with skull base tumors suffer from symptoms which are directly related to their disease. Table 3 lists some of the common symptoms associated with tumors originated in the skull base and with the postoperative morbidity frequently encountered following their extirpation. The anterior skull base surgery (ASBS) questionnaire (Table 2) was established in 2003 as the main instrument for evaluating QOL in patients with anterior skull base neoTable 2
plasms (20). The ASBS questionnaire is a disease-specific instrument based on a pooling of questions related to general QOL (SF-36; Glasgow Benefit Inventory; The Center for Epidemiologic Studies Depression Scale) and to head and neck cancer (EORTC QLQ-C30 and UW-QOL). Questions were added to this instrument in order to cover symptoms specifically associated with anterior skull base tumors and postoperative morbidity (Table 3). These questions were generated from interviews with health professionals and surgeons as well as from interviews with patients and their caregivers. They covered problems with various aspects of taste, smell, appearance, epiphora, nasal secretions, and visual disturbances, which are frequently found in this patient population. During the development of this instrument, patients were asked to rate each item for its relative importance to their own QOL. They were also asked to contribute questions that they considered as being important but that had not been included in the questionnaire. Statistical analysis identified 35 items divided into six domains of QOL: role of performance (six items), physical function (seven items), vitality (seven items), pain (three
Disease-Specific Instruments
Name
Description
Advantage
ASBS
The only disease-specific instrument developed especially for patients with tumors of the anterior skull base. A self-administered instrument with 35 items divided into 6 domains. A self-administered instrument with 21 items developed for patients with HN cancer. Based on the QLQ-C30 generic instrument. A disease-specific, self-administered instrument developed for HN cancer patients. Based on the FACT-G generic instrument. A disease-specific, self-administered instrument developed for HN cancer patients. Version 4 also includes psychological domains. A disease-specific instrument with 21 items and 4 domains. Interviewer administered
Tested for validity, reliability, and internal consistency. Available in English, Hebrew, and Portuguese. Tested for validity and reliability in HN patients.
Validated only for patients with anterior skull base tumors.
Can be coadministered with FACT-G, fast application.
Not independently validated. Not validated for SB patients.
Reliability and validity in study.
Limited coverage, useful if combined with other measures.
Reliability and validity in study.
Further validation required.
EORTC HN Module
FACT-HN
UWQOL
University of Michigan HN QOL
Disadvantage
Not validated for SB patients.
Abbreviations: EORTC, QOLQ-C30, European organization for research and treatment of cancer and quality-of-life questionnaire; FACT-G, functional assessment of cancer therapy, general module; ASBS, anterior skull base surgery; HN, head and neck; SB, skull base; UWQOL, University of Washington Quality of Life Scale.
Chapter 13: Quality of Life of Patients with Skull Base Tumors
191
Table 3 Functional Disabilities as a Result of Skull Base Tumors or its Extirpation and its Effect on Quality of Life Domains Organ involved
Functional disability
Vestibulocochlear nerve Olfactory nerve ablation Optic nerve and chiasma Orbital disfunction or exenteration Facial nerve Vagus nerve Glossopharyngeal and hypoglossal nerves Maxillofacial bones Trigeminal nerve Meninges Brain parenchyma
Hearing, communication Change in appetite, olfactory and gustatory function Visual Visual
Social function, role of performance Physical functioning Physical functioning Physical functioning
Cosmetic disfigurement and visual dysfunction Speech and swallowing, aspirations, and weight loss Swallowing, aspirations, and weight loss
Social function, psychological and physical functioning Social function, physical functioning Physical functioning
Cosmetic deformity, speech and swallowing Paraesthesia, neuralgia, use of painkillers CSF leak, recurrent meningitis Personality changes, mental impairment
Social function, psychological and physical functioning Pain, psychological functioning, role of performance Pain, psychological functioning, role of performance Social function, psychological and physical functioning, role of performance
items), influence upon emotions (five items), and specific symptoms (seven items). Table 4 summarizes the different QOL domains and items included in the ASBS questionnaire. An established approach for validating an instrument of this type is to examine its construct validity, which means testing whether specific clinical variables (i.e., clinical measures of disease severity) agree with the QOL domain score as hypothesized. Alternatively, construct validity estimates the ability of the instrument to differentiate between patient groups considered as having different health status. For example, it is expected that older patients will show significantly poorer scores relative to younger patients in physical function domains. Similarly, patients with malignant lesions undergoing adjuvant radiation therapy may have poorer scores than patients with benign lesions. Construct validity of the ASBS questionnaire was assessed by testing whether the clinical variable of the patient did influence the QOL domain score as hypothesized. The results showed that patients >60 years of age had signifi-
Table 4
Affected QOL domain
cantly poorer scores in the role of performance and physical function domains than younger patients. Also, patients with malignant tumors had significantly poorer scores in the domains of specific symptoms, influence upon emotions, physical function, and role of performance compared to patients with benign tumors (Fig. 1). Radiotherapy was associated with poorer scores in the domains of specific symptoms and influence upon emotions, and comorbidity was associated with poor physical function scores. The sensitivity of the instrument to changes over time was tested by comparing the scores of patients operated 3–6, 6–24, and >24 months before the study was activated. Using this approach, a gradual improvement in QOL measures was found during the first 6–24 months after surgery. A similar change in QOL measures after skull base surgery was found by De Jesus et al. (19) for meningiomas involving the cavernous sinus. A recent prospective study further established the ASBS questionnaire as the main instrument for evaluating QOL of patients with anterior skull base tumors
Studies Estimating Quality of Life (QOL) of Patients with Skull Base Tumors
Study
Tumor
Questionnaire
Gil et al. (20)
ASB
ASBS
Fukuda et al. (21)
ASB
NA
Mohsenipour et al. (22)
Meningioma
IHD and NHP
Lang et al. (23)
Meningioma
GOS, SF-36
Nikolopoulos et al. (18) Tufarelli et al. (24)
AS AS
GBI SF-36 and DHI
de Cruz et al. (25) Baumann et al. (26) Irving et al. (16) Myrseth et al. (11)
AS AS AS AS
SF-36 SF-36 EORTC, QOLQ-C30 GBI and SF-36
Sandooram et al. (27)
AS
GBI
Martin et al. (28)
AS
SF-36
Main finding 38% had a significant improvement and 36% reported no change in QOL after surgery 89% had some complaints and 63% reported worsened QOL after surgery 60% had mild-to-moderate impairment and 20% had moderate-to-severe impairment of their QOL 76% had good-to-moderate recovery 1 yr after surgery, but most of the patients were still functioning below accepted norms >53% of patients reported an overall decline in QOL QOL scores were lower than in the general population. Significant deterioration in elderly patients. Reduced QOL scores relative to a matched healthy population. Significant reductions in QOL compared to norms. Excellent QOL after resection of acoustic schwannomas 1.5 cm. QOL significantly better in the radiosurgery group than in the microsurgery group. Conservative management was associated with a better QOL compared with radiosurgery or microsurgery QOL was rated significantly below published norms. The physical functioning, vitality, and social functioning were reduced.
Abbreviations: ASBS, anterior skull base surgery; AS, acoustic schwannoma; NA, not applicable; IHD(NS), Innsbruck Health Dimensions Questionnaire for Neurosurgical Patients; GBI, Glasgow benefit inventory; DHI, Dizziness Handicap Inventory; NHP, Nottingham Health Profile; GOS, Glasgow Outcome Score; EORTC, QOLQ-C30, European organization for research and treatment of cancer quality of life questionnaire.
192
Gil and Fliss
Age
Radiotherapy
Role of performance
A
3.0
Role of performance
B
>60
Yes
3.0
<60
No
2.8
2.2
2. 4
2. 2
2. 2 2.4
2.4
2.8
2.6
2.6 2 .8 3 .0
3.0
Vitality
2.6
2 .8
3.0
2 .6
2. 8
2.4
2. 2
2.2
2 .4
2. 6
2.4
3.0
Specific symptoms
2.4
2. 8
2.6
2. 2
2. 6
2.8
2. 2
2 .4
3 .0
2. 4
2.4
2.2
2. 2
Vitality
2.6 2.8
2.8
3.0
3.0
Pain
Pain
Type of surgery
Comorbidity
Role of performance
C
Role of performance
D
Wide resection
3.0
Yes
3.0
Limited resection
No
2.8
2.8
Vitality
2. 8 2.6 2.4
2.2 2.4 2. 6
2.4
3 .0
2.8
2.6
2.2
2 .2
2.6
2 .8
3. 0
3 .0
2.8
2.4
2.2 2.4 2.6
2 .2
2.6
2. 4
2.4
3 .0
2.6
2. 2
2. 8
Specific symptoms
2.2
2. 8
2.4
2. 6
Vitality
Physical function
2.2
2.2
2.4
2. 8
2.4
2. 2
3 .0
2.6 2.6
3.0
2 .8
2.4
Specific symptoms
Role emotional
2 .2
2.2
2. 6
2.4
Physical function
3. 0
3 .0
2.6
2.8
Role emotional
Physical function
2 .6
2.4
2.2
3 .0
2.6
2 .2
Specific symptoms
Role emotional
2. 8
2. 8 2. 6
2.4 2.2
Physical function
3.0
2.6
3. 0
Role emotional
2.8
2.8 3.0
3.0
Pain
Pain
Figure 1 Polar graphs of QOL questionnaire scores distributed to demographic and clinical variants. Each graph (A–D) shows the average scores of six specific QOL domains represented by each axis. A higher score indicates a better QOL. The effect of (A) age (below or above 60 years), (B) radiotherapy, (C) type of surgery, and (D) patients’ comorbidity on different QOL measures are shown. Source: From Ref. 20.
and demonstrated its ability to distinguish between different patients at a given point of time from surgery (20). The complete questionnaire is found in the Appendix of this chapter. Health-related QOL questionnaires can be selfadministered or completed with the assistance of a trained interviewer. Although an interviewer-administered mode is resource intensive, it ensures compliance and decreases errors and the number of missing items due to misunderstanding.
QUALITY OF LIFE OF PATIENTS WITH SKULL BASE TUMORS Patients with Anterior Skull Base Tumors Surgical procedures for resection of neoplasms of the skull base can carry significant morbidity. Previous studies have shown that up to 44% of these patients may suffer from anosmia as a result of tumor resection (6). Other significant complications that may influence the functional outcome of patients are hearing loss, meningitis, cerebrospinal fluid leak,
Chapter 13: Quality of Life of Patients with Skull Base Tumors
Overall QOL
50
A
40 30
B
40 30 20
10
10
0
0 Worse
Same
Worse
Better
Financial status
60
C
Same
Better
Impact upon emotions
60
D
Patients (%)
Patients (%)
50
20
50
Social activity
60
Patients (%)
Patients (%)
60
193
40 30 20
50 40 30 20
10
10
0 Worse
Same
Better
Worse
Same
Better
Figure 2 The effect of surgery on different quality-of-life measures. (A) Overall quality of life. (B) Social activity. (C) Financial status. (D) Emotional state. Source: From Ref. 20.
osteoradionecrosis, fistula, mucocele, and visual disturbances (Table 3). In an early study, Janecka et al. used the Karnofsky performance score to evaluate the effect of cranial base surgery on patients’ QOL (20). They showed that 83% of these patients had improved or unchanged scores after surgery. Recently, the long-term QOL of patients undergoing anterior skull base tumor resection was assessed using a diseasespecific instrument (20,29–31). Figure 2 displays the mean change in QOL after anterior skull base tumor extirpation. The results found by Gil et al. showed that 38% of the patients reported a significant improvement in overall QOL and that 36% reported no change in QOL after surgical removal of their tumor (20). The minority of patients (26%) responded that the surgical procedure worsened their QOL. The scores of the social, financial, and emotional state domains were lower relative to the overall QOL score. The mean physical function score of patients younger than 60 years was significantly higher than that of older patients (Fig. 1). Patients with benign tumors reported higher scores in the domains of role of performance, physical function, specific symptoms, and impact upon emotions than patients with malignant tumors. Interestingly, both groups reported only a minor influence of the aspect of pain on their QOL. Perioperative radiotherapy was associated with significantly lower scores in the domains of specific symptoms and impact upon emotions. Extensive cranial resection was also associated with lower scores in the impact upon emotions domain. As expected, patients with an additional illness reported lower scores of QOL in the physical function domain.
Other aspects of QOL are the length of recovery after surgery and the stability of scores over time. The retrieved data showed improvement in QOL measures at 6 to 24 months following surgery compared to the 3 to 6 months following surgery and stability of the scores occurring >6 months postoperatively (Fig. 3). The overall results of this study showed that resections of tumors of the anterior skull base have a positive impact on the patients’ QOL. The relatively good scores recorded for the role of performance, pain, and specific symptoms domains further demonstrate this positive impact of surgery on different aspects of QOL. Another study on patients with anterior skull base tumors was performed by Fukuda et al. (21). In their study, the authors evaluated 13 patients who underwent classical craniofacial resection of malignant tumors: 12 patients had some complaints, and 8 of them reported that their QOL worsened after surgery. Mohsenipour et al. (22) recently showed that the majority of patients after meningioma resection had mild-tomoderate impairment of QOL, and that there was severe impairment in the physical handicap and energy level domains. Similar results were reported for meningiomas involving the cavernous sinus (19). In another study that looked at the outcome of 105 patients after excision of petroclival meningioma in which a standard lateral skull base approach was used (23), 76% of the patients reported good-to-moderate recovery at 1 year postoperatively, but 39–72% of these patients were still functioning below accepted norms.
Mean QOL score
194
Gil and Fliss
4.0 3.5 3.0 2.5 2.0
0.0
6.0
12.0
Time after surgery (months) Figure 3 Estimating the recovery period and stability of QOL measures before and after anterior skull base tumor resection. The average score of the overall health-related QOL before surgery and at 6 and 12 months afterward is shown. Bars indicate ±SEM.
Studies on patients with malignant intracranial tumors have demonstrated severely deteriorating QOL measures, with marked decline in cognitive, physical, emotional, and social functioning after surgery (32). The stability of QOL following surgery is an important issue of patients’ QOL and recovery. A dynamic evaluation of QOL after ASBS and excision of meningiomas involving the cavernous sinus showed a gradual improvement of QOL measures during the first year after surgery and stability of scores during the second year (33) Gradual deterioration of QOL, however, was observed during the postoperative period in patients with intracranial malignancy. Old age was associated with a negative impact on health-related QOL, following resection of anterior skull base tumors or cavernous sinus meningiomas (19,20). In contrast, young patients after excision of acoustic neuromas had a worse overall QOL and financial status compared to old ones (18). The discrepancy between these pathologies may stem from the fact that young patients with acoustic neuromas had a good preoperative QOL compared to young patients with anterior skull base tumors who invariably suffered from serious symptoms prior to surgery. It can be also presumed that younger patients cope better with anterior skull base neoplasms and surgery than do older patients with similar conditions. Perioperative radiotherapy was also found to predict poor QOL scores, especially in the specific symptoms and emotional functioning domains. Radiation therapy is used for patients with malignant tumors, and both factors in combination probably contributed to the deteriorating emotional states of these patients. Similar impacts on the patient’s overall QOL and emotional state were found in patients receiving radiation therapy for head and neck cancer (34) or other neoplasms (35).
Patients with Acoustic Schwannoma Estimation of health-related QOL is particularly important in patients with acoustic schwannoma. Nikolopoulos et al. (18) used the Glasgow Benefit Inventory for evaluating the patient’s QOL after resection of acoustic schwannoma and showed that more than 53% of their 53 patients reported an overall decline in QOL. They also showed that 29% of
their patients had declined financial status as a consequence of the operation. Tufarelli et al. (24) used both the SF-36 and the Dizziness Handicap Inventory and found significant deterioration of QOL among females and patients >45 years. Similar results were reported by de Cruz et al. (25) for the translabyrinthine and retrosigmoid approaches and by Baumann et al. (26) for the middle cranial fossa approach. In contrast, Irving et al. (16) used the EORTC generic instrument to show an excellent QOL after resection of acoustic schwannomas, especially for tumors smaller than 1.5 cm. An interesting study performed by Myrseth et al. (11) evaluated the QOL of patients treated for unilateral vestibular schwannoma either by microsurgery or by gamma knife radiosurgery. The QOL, posttreatment facial nerve function, and hearing were significantly better in the radiosurgery group. Improved QOL scores have been associated with better facial nerve and hearing function. In contrast, Sandooram et al. (27) suggested that a conservative management approach is associated with a better QOL outcome compared with both radiosurgery and microsurgery. In the largest study performed to date, Ryzenman et al. (36) used a disease-specific questionnaire and demonstrated that low QOL scores were specifically associated with postoperative headache symptoms after excision of schwannomas. Similar result was reported by Betchen et al. (37) using generic instruments. A negative impact of surgery on patients with acoustic schwannomas is expected since the preoperative healthrelated QOL is hardly affected by their clinical condition, especially when the tumors are small. In contrast, patients with anterior skull base tumors present with major debilitating symptoms, such as nasal obstruction and secretion, diplopia, epiphora, epistaxis, headache, visual disturbances, and facial deformities. Most of these symptoms are relieved following tumor extirpation, so it can be expected that the patients will experience some degree of improvement in their QOL following surgery. Table 5 summarizes the published data on the QOL of patients with skull base tumors of the anterior or lateral fossa.
Assessment of Quality of Life by Proxy Evaluation of QOL involves retrieving information on the physical, emotional, social, and economical aspects of patients’ activities of daily living, as well as on specific symptoms associated with their disease. Who should measure the QOL of patients? The primary source of information on a patient’s QOL is undoubtedly the patient himself. Proxy assessment of QOL may, however, be important as an adjunct to patients’ rating for several reasons: (1) it improves compliance and allows more accurate evaluation of patients’ conditions, especially for those whose medical conditions are poor; (2) it allows proper evaluation of patients with neuropsychologic dysfunction, who may be incapable of providing QOL information themselves; and (3) it serves as complementary data to increase score reliability and prevent potential bias. As mentioned earlier, previous works had focused on the subjective perception of QOL by the patient alone, but lay caregivers and health professionals can also play an important role in the rehabilitation and day-to-day care of patients both before and after surgery. The perception of a patient’s condition by the surgeon and caregiver is particularly important following complex surgeries that bear significant morbidity for which multidisciplinary medical treatment and rehabilitation efforts are required for long periods of time. The evaluation of QOL in patients with anterior skull base tumors is particularly challenging for several reasons.
Chapter 13: Quality of Life of Patients with Skull Base Tumors Table 5
Summary of Domain Questions in the ASBS QOL Questionnaire
Role of performance General performance Performance at work Performance at home Participation in social activities Communication with people Effect of health on performance Physical functioning Climbing stairs Leaning and standing Walking long distances Walking short distances Preferring to stay in bed Carrying out routine activities Reducing extent of physical activity because of disease Vitality Feeling weak Feeling tired Achieving goals Feeling enthusiastic Feeling motivated Feeling energetic Relationships with spouse Pain Experiencing pain Effect of pain on activity Use of painkillers Specific symptoms Change in appetite Altered sense of taste Altered sense of smell Altered appearance Nasal secretions Eye secretions Affect of surgery on vision Impact upon emotions Feeling tense Problems falling asleep Feeling worried or frustrated Feeling relaxed and calm Worrying about finances
195
showed a trend to overrate patients’ QOL, and only caregivers of patients over 60 years of age reported scores similar to that of the patient. Interestingly, patients’ with recurrent disease had the strongest agreement with their caregivers (r = 0.94). The general cross−correlation between caregivers and patients reporting is displayed in Figures 4 and 5. While the caregiver reported similar scores as the patient, the operating surgeon tended to overrate his patients’ QOL. Gil et al. showed that the surgeon’s perception of his patient’s QOL is not sufficiently accurate to correctly estimate the patients’ true QOL (30). The authors concluded that these judgments should come from the patient or his caregiver. Similar results were reported on outpatient oncological patients (40), patients with head and neck cancer (14), and patients with malignant brain tumors (41). Examination of the surgeon–patient agreement at both the group and the individual levels revealed overrating of patients’ QOL by the surgeon (30). This finding confirmed those of other studies that examined whether a physician or a health professional can make a valid assessment of his patients’ QOL (39,42). Similar studies on the perception of cancer-related fatigue also showed close agreement between patient and caregiver scoring, whereas there was a weaker correlation between patient and oncologist rating (43). In essence, the surgeon apparently believes that the patient has a better QOL than what the patient himself perceives. The surgeon’s perception of patients’ QOL is not sufficiently accurate for judging the physical, emotional, social, and economical status of his patients, and this suggests that health-care professionals need to work with patients and caregivers simultaneously to facilitate communication and discussion about expectations and choice of medical treatment. An original study by Lang et al. (29) estimated the influence of transpetrosal excision of petroclival meningiomas on the QOL of the patient’s caregiver. They found that more than one-half of the caregivers reported negative change in their lifestyle and more than one-third experienced a major decline in their work. They concluded that skull base procedures can not only affect the patient’s QOL but that they can also significantly increase the burden on the patient’s caregiver.
How Can We Improve Patients’ QOL? First, many of these patients have intracranial extension of their tumor (i.e., the patients are prone to brain injury during surgery and subsequently to cognitive dysfunction). Second, their medical treatment requires extensive surgical intervention, prolonged hospitalization, and long period of rehabilitation. Finally, many of these patients require multiple surgeries. All these factors taken together may eventually lead to problems with the accuracy of patients’ self-reporting. Little is known about the perceptions of QOL following a major head and neck surgery among cancer patients, their families, and the operating surgeon. The prevailing opinion of the limited ability of caregivers and oncologists to correctly assess patients’ QOL was challenged by recent studies, which found a good agreement between patient and proxy rating (30,38,39), that performed a triple survey in an effort to elucidate the lay caregivers’ and surgeons’ perception of patients’ QOL. An overall significant agreement was found between patients’ and caregivers’ scores, both at the group level (mean scores of each domain) and individual level (each patient’s pair). There was a minor correlation in the influence upon emotions domain and no correlation in the pain domain. Interestingly, they found no correlation between the surgeons’ and the patients’ ratings. In general, the caregivers
As physicians, we are trained to give our patients optimal treatment for achieving the highest survival rate with the least possible morbidity and mortality. Managing patients’ QOL is often forgotten in the shadow of providing the possibility for a cure. While increasing efforts have been made during the last 100 years to develop new surgical approaches and novel adjuvant modalities for the treatment of cancer, little progress has been made in developing new techniques for improving patients’ QOL. Extensive procedures involving extirpation of skull base tumors may be associated with high morbidity rates, thus having a negative impact on QOL. Earlier in this chapter, a review of the published data on QOL of patients with skull base tumors revealed that most patients with anterior skull base neoplasms show similar or better QOL scores following surgery with or without radiation. In contrast, most patients with acoustic schwannomas show deterioration of their QOL as a result of any clinical intervention (i.e., surgery or radiosurgery).
Control of Pain Few papers have assessed the level of chronic pain in head and neck cancer patients (44–49). In a recent study, the level of pain in 100 patients scheduled for oncological head and neck surgery was assessed prior to their operation, and only 10%
196
Gil and Fliss
Patient Caregiver 3.0
Mean score
2.8
2.6
2.4
2.2
2.0 >60 Malignant Benign No
Average <60
Age
Tumor
Yes
Radiation
No
Yes
Comorbidity
Figure 4 Demographic and clinical influence on patient-proxy scores. Mean scores of patients’ ratings (red bars) and caregivers’ ratings (green bars) of patients’ QOL. In general, caregivers reported higher scores than the patients themselves.
Specific symptoms 4.5 4.0 3.5
Health giver score
3.0 2.5 2.0 1.5 r = 0.71 1.0
p < 0.0001
0.5 0.0 0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Patient score
Figure 5 Relationship between patients’ ratings (x-axis) and caregivers’ ratings (y-axis) of patients’ health-related quality of life (QOL). Each point represents one patient–caregiver pair. Some of the points overlap. The graph shows the mean score of all QOL domains. The correlation coefficient r and statistical significance p are marked in the right bottom corner of each graph. The red line is the best-fit straight line to the measured points. The dashed green lines depict the 95% confidence intervals. Source: From Ref. 30.
of them reported that they were suffering from a significant level of pain as a result of their disease (50). One-half of the patients who were suffering from significant levels of pain had craniofacial tumors. Chaplin et al. (44) also reported that 8% of their patients suffered from severe pain prior to surgery and noted that the highest level of pain was among patients suffering from tumors originating in the nasopharynx. How can we improve patients’ QOL after resection of skull base tumors? One way to positively affect their postoperative QOL involves pain control. Skull base procedures are associated with a considerable level of pain during the early postoperative period (51). Numerous studies have assessed pain issues in oncological patients during the last decade (44), but the levels of pain following head and neck surgery in general and craniofacial or skull base procedures in particular have not been systematically evaluated. While many analgesic regimens are available [e.g., pro re nata (PRN), scheduled dosing, intravenous patient-controlled intravenous analgesia , little is known about the optimal therapeutic modality for the treatment of acute postoperative pain after head and neck procedures. In a multidepartmental survey of acute pain level measured following surgery, the neurosurgicaland craniofacial-related pain was rated 5 and 7 of 11, respectively, coming after orthopedic, abdominal, gynecologic, and urologic procedures and preceding plastic, ophthalmologic, vascular, and oral procedures (51). In a recent study, Gil et al. (50) estimated the level of pain following skull base surgery and used their data to tailor a specific pain-control protocol specific to each procedure. The overall results of that study showed that a PRN protocol is not adequate for the management of pain following skull base and maxillofacial procedures. Based on their findings, the authors suggest treating patients undergoing maxillofacial procedures with IV-PCA during days 1 to 3 and then with scheduled nonsteroidal anti-inflammatory drugs (NSAIDs), regardless of whether the patient complains of pain. The new regimen
Chapter 13: Quality of Life of Patients with Skull Base Tumors
showed a sharp reduction in the level of pain during the postoperative period among this subpopulation of patients. Patients scheduled for skull base procedures also suffered a moderate degree of pain shortly after surgery, unlike patients undergoing maxillofacial osteotomies; however, these patients reported a slow reduction in the level of pain (t0.5 = 6 days). Patients undergoing skull base procedures most frequently required continuous drainage of cerebrospinal fluid due to violation of the dura during surgery, and this mandates bed rest for 5 days until the drain is removed (52). These patients complained mainly of musculoskeletal and back pain as a result of lying in a supine position for a prolonged period of time. The authors recommended NSAIDs for these patients whether or not they requested painkillers. Patients after intracranial and skull base procedures are at risk for developing severe intracranial complications, such as meningitis, tension pneumocephalus, and hemorrhage. Treatment of pain with a non-opiate analgesic allows close monitoring of the patient’s neurological and mental conditions. It was therefore preferred to administer tramadol PRN as an adjuvant pain control, since this drug does not cause much sedation. NSAIDs have analgesic properties that are associated with opioid-sparing options and few side effects, and adequate postoperative analgesia may be provided by the routine administration of these drugs (53). Furthermore, because these drugs do not induce respiratory depression or sedation, their administration may facilitate less intensive postoperative monitoring and allow close neurological follow-up after intracranial procedures. In addition, NSAIDs do not require sophisticated delivery mechanisms. If administered at regular intervals, regardless of whether or not the patient requested pain relievers, NSAIDs can decrease patient dependence on opioids with satisfactory results.
A Multidisciplinary Team Approach In an effort to provide effective and efficient care to patients with chronic health conditions, the U.S. health-care system is in the process of redesigning its delivery system. One approach to meet the high demands of patients and to best utilize resources is the use of a multidisciplinary team approach to provide better care, compared to an individual patient–physician care protocol. When properly implemented, this team approach provides positive measurable outcomes. With a diverse group of health-care professionals, such as physicians, nurses, pharmacists, dietitians, and health educators with the patient at the center, the team can ensure that treatment goals are maintained for chronic diseases. The team approach can enhance patient satisfaction and self-management, development of a community support network, team coordination, team communication, patient follow-up, use of protocols and other tools, use of computerized information systems, and outcome evaluations. A multidisciplinary team can provide more comprehensive care than an individual patient–surgeon approach and so enhance the patient’s QOL. The management scheme should also include patient satisfaction and QOL questionnaires, self-management brochures, employment of a community support network, team coordination, team communication, scheduled follow-up, use of database information systems, and outcome measures. Paramedical support should involve practical, informational, and emotional support. Screening patients for psychosocial distress can be conducted quickly and is important to identify patients requiring additional interventions. Based on previous reports, the most influential factor on the QOL of patients with skull base tumors was malignancy, a factor that led to a signifi-
197
cant decrease in the overall score (20). Radiotherapy, old age, comorbidity, and wide resection also significantly worsened QOL scores of specific domains (20). The main effort to improve QOL should focus on this subpopulation of patients who are prone to suffer significant impairment of their QOL. Psychosocial barriers and family issues are also key factors in determining QOL. The role of the health-care professionals who are part of the multidisciplinary team involves not only providing medical follow-up and treating short-term complications as a result of surgical or adjuvant therapy, but also supervising appropriate monitoring of long-term complications for early detection of depression or financial problems. Sources of assistance can be accessed by telephone, e-mail, and dedicated web pages, and they can include support groups, specialists who answer questions online and weblogs. (blogs).
Psychotherapy, Antidepressants and Group Support Up to two-thirds of patients with cancer may suffer from depression, with half of them having at least one episode of major depression (54). Pharmacological treatment was shown to induce significant improvement in depression symptoms compared to baseline within 6 weeks after administration antidepressant therapy (55,56). Psychotherapeutic cognitive and behavioral therapies were found to have an impact on the level of depression and symptoms in patients with cancer, and combined behavioral and pharmacological therapies were shown to even further reduce the level of depression in these patients (57,58). Psychological intervention by means of group support has been shown not only to increase QOL but also to provide survival benefit for patients with metastatic carcinoma of breast (59). Other studies have established the benefit of psychological group support in patients with other malignancies, including malignant melanoma, hematologic malignancies, and gastrointestinal cancers (60,61) In contrast, one recent randomized trial of group support failed to show any impact on health-related QOL (62). A module of psychological group support can include 5 to 10 patients with similar disease conditions and two leaders. The leaders should include one member of the surgical team and one member of the nursing team, along with a psychiatrist, a psychologist, or a social worker who are experienced in leading group support therapy (63). Group support may include discussions on relationships with family, friends, and the health caregivers; the effect of the disease on patient’s dayto-day life; and past/future treatment modalities. Another way of decreasing patients’ distress is by introduction to and discussion with cancer survivors who had been treated by the same multidisciplinary team. Patients may also be trained to practice relaxation exercises during each seminar and at home with or without the aid of a spouse or health caregiver. Finally, a web-based social support group can also be effective in reducing depressive symptoms in cancer patients (64).
CONCLUSIONS The overall QOL in the majority of patients after anterior skull base tumor extirpation can be classified as “good,” with significant improvement taking place within 6 months following surgery. Patients with acoustic schwannomas suffer from significant deterioration in their QOL after any intervention, including both surgery and radiosurgery. Data retrieved from specially designed questionnaires revealed that the financial and emotional QOL domains had the worse impact on
198
Gil and Fliss
the patients. Old age, malignancy, comorbidity, radiotherapy, and extensive surgery were found to be negative prognostic factors for QOL measures. Pain control regimens, antidepressants and other psychological modalities, including group support, can improve QOL measures in these patients. A multidisciplinary team approach emerged as being the best intervention modality for enhancing the health-related and overall QOL of patients with skull base tumors.
REFERENCES 1. Portenoy RK. Quality of life issues in patients with head and neck cancer. In: Harrison L, ed. Head and Neck Cancer. New York: Plenum Press,1995:218–231. 2. Gliklich RE, Goldsmith TA, Funk GF. Are head and neck specific quality of life measures necessary? Head Neck. 1997;19:474– 480. 3. Boyle JO, Shah KC, Shah JP. Craniofacial resection for malignant neoplasms of the skull base: An overview. J Surg Oncol. 1998;69:275–284. 4. Irish JC, Gullane PJ, Gentili F, et al. Tumors of the skull base: Outcome and survival analysis of 77 cases. Head Neck. 1994;16:3– 10. 5. Raveh J, Laedrach K, Speiser M, et al. The subcranial approach for fronto-orbital and anteroposterior skull-base tumors. Arch Otolaryngol Head Neck Surg. 1993;119:385–393. 6. Fliss DM, Gil Z, Spektor S, et al. Skull base reconstruction after anterior subcranial tumor resection. Neurosurg Focus. 2002;12:1– 7. 7. Hassan SJ, Weymuller EA Jr. Assessment of quality of life in head and neck cancer patients. Head Neck. 1993;15:485–496. 8. McHorney CA, Ware JE Jr, Raczek AE. The MOS 36-Item ShortForm Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care. 1993;31:247–263. 9. Tufarelli D, Meli A, Alesii A, et al. Quality of life after acoustic neuroma surgery. Otol Neurotol. 2006;27:403–409. 10. Baumann I, Polligkeit J, Blumenstock G, et al. Quality of life after unilateral acoustic neuroma surgery via middle cranial fossa approach. Acta Otolaryngol. 2005;125:585–591. 11. Myrseth E, Moller P, Pedersen PH, et al. Vestibular schwannomas: Clinical results and quality of life after microsurgery or gamma knife radiosurgery. Neurosurgery. 2005;56:927–935. 12. Betchen SA, Walsh J, Post KD. Self-assessed quality of life after acoustic neuroma surgery. J Neurosurg. 2003;99:818–823. 13. Kelleher MO, Fernandes MF, Sim DW, et al. Health-related quality of life in patients with skull base tumours. Br J Neurosurg. 2002;16:16–20. 14. Ware J Jr, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: Construction of scales and preliminary tests of reliability and validity. Med Care. 1996;34(3):220–233. 15. Aaronson NK, Ahmedzai S, Bergman B, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: A quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst. 1993;85:365–376. 16. Irving RM, Beynon GJ, Viani L, et al. The patient’s perspective after vestibular schwannoma removal: Quality of life and implications for management. Am J Otol. 1995;16:331–337. 17. Singh B, Bhaya M, Stern J, et al. Validation of the Charlson comorbidity index in patients with head and neck cancer: A multiinstitutional study. Laryngoscope. 1997;107:1469–1475. 18. Nikolopoulos TP, Johnson I, O’Donoghue GM. Quality of life after acoustic neuroma surgery. Laryngoscope. 1998;108:1382– 1385. 19. De Jesus O, Sekhar LN, Parikh HK, et al. Long-term followup of patients with meningiomas involving the cavernous sinus: Recurrence, progression, and quality of life. Neurosurgery. 1996;39:915–919. 20. Gil Z, Abergel A, Spektor S, et al. Quality of life following surgery for anterior skull base tumors. Arch Otolaryngol Head Neck Surg. 2003;129:1303–1309.
21. Fukuda K, Saeki N, Mine S, et al. Evaluation of outcome and QOL in patients with craniofacial resection for malignant tumors involving the anterior skull base. Neurol Res. 2000;22:545–550. 22. Mohsenipour I, Deusch E, Gabl M, et al. Quality of life in patients after meningioma resection. Acta Neurochir (Wien). 2001;143:547–553. 23. Lang DA, Neil-Dwyer G, Garfield J. Outcome after complex neurosurgery: The caregiver’s burden is forgotten. J Neurosurg. 1999;91:359–363. 24. Tufarelli D, Meli A, Alesii A, et al. Quality of life after acoustic neuroma surgery. Otol Neurotol. 2006;27:403–409. 25. da Cruz MJ, Moffat DA, Hardy DG. Postoperative quality of life in vestibular schwannoma patients measured by the SF36 Health Questionnaire. Laryngoscope. 2000;110:151–155. 26. Baumann I, Polligkeit J, Blumenstock G, et al. Quality of life after unilateral acoustic neuroma surgery via middle cranial fossa approach. Acta Otolaryngol. 2005;125:585–591. 27. Sandooram D, Grunfeld EA, McKinney C, et al. Quality of life following microsurgery, radiosurgery and conservative management for unilateral vestibular schwannoma. Clin Otolaryngol Allied Sci. 2004;29:621–627. 28. Martin HC, Sethi J, Lang D, et al. Patient-assessed outcomes after excision of acoustic neuroma: Postoperative symptoms and quality of life. J Neurosurg. 2001;94:211–216. 29. Abergel A, Gil Z, Spektor S, et al. Quality of life following anterior skull base surgery. Harefuah. 2004;143:489–493. 30. Gil Z, Abergel A, Spektor S, et al. Patient, caregiver, and surgeon perceptions of quality of life following anterior skull base surgery. Arch Otolaryngol Head Neck Surg. 2004;130:1276–1281. 31. Gil Z, Abergel A, Spektor S, et al. Development of a cancerspecific anterior skull base quality-of-life questionnaire. J Neurosurg. 2004;100(5):813–819. 32. Osoba D, Aaronson NK, Muller M, et al. Effect of neurological dysfunction on health-related quality of life in patients with highgrade glioma. J Neurooncol. 1997;34:263–278. 33. Gil Z, Abergel A, Fliss DM. Quality of Life in Patients with Anterior Skull Base Tumors: A prospective study. Phoenix, AZ: North American Skull Base Society, 2006. 34. Shah S, Har-El G, Rosenfeld RM. Short-term and long-term quality of life after neck dissection. Head Neck. 2001;23:954–961. 35. Chandra PS, Chaturvedi SK, Channabasavanna SM, et al. Psychological well-being among cancer patients receiving radiotherapy–a prospective study. Qual Life Res. 1998;7:495–500. 36. Ryzenman JM, Pensak ML, Tew JM Jr. Headache: A quality of life analysis in a cohort of 1,657 patients undergoing acoustic neuroma surgery, results from the acoustic neuroma association. Laryngoscope. 2005;115:703–711. 37. Betchen SA, Walsh J, Post KD. Self-assessed quality of life after acoustic neuroma surgery. J Neurosurg. 2003;99:818–823. 38. Sneeuw KC, Aaronson NK, Osoba D, et al. The use of significant others as proxy raters of the quality of life of patients with brain cancer. Med Care. 1997;35:490–506. 39. Deschler DG, Walsh KA, Friedman S, et al. Quality of life assessment in patients undergoing head and neck surgery as evaluated by lay caregivers. Laryngoscope. 1999;109:42–46. 40. Kurtz ME, Kurtz JC, Given CC, et al. Concordance of cancer patient and caregiver symptom reports. Cancer Pract. 1996;4:185– 190. 41. Hahn CA, Dunn RH, Logue PE, et al. Prospective study of neuropsychologic testing and quality-of-life assessment of adults with primary malignant brain tumors. Int J Radiat Oncol Biol Phys. 2003;55:992–999. 42. Yeager KA, Miaskowski C, Dibble SL, et al. Differences in pain knowledge and perception of the pain experience between outpatients with cancer and their family caregivers. Oncol Nurs Forum. 1995;22:1235–1241. 43. Vogelzang NJ, Breitbart W, Cella D, et al. Patient, caregiver, and oncologist perceptions of cancer-related fatigue: Results of a tripart assessment survey. The Fatigue Coalition. Semin Hematol. 1997;34:4–12. 44. Chaplin JM, Morton RP. A prospective, longitudinal study of pain in head and neck cancer patients. Head Neck. 1999;21:531– 537.
Chapter 13: Quality of Life of Patients with Skull Base Tumors 45. Olsen KD, Creagan ET. Pain management in advanced carcinoma of the head and neck. Am J Otolaryngol. 1991;12:154–160. 46. Vecht CJ, Hoff AM, Kansen PJ, et al. Types and causes of pain in cancer of the head and neck. Cancer. 1992;70:178–184. 47. Grond S, Zech D, Lynch J, et al. Validation of World Health Organization guidelines for pain relief in head and neck cancer. A prospective study. Ann Otol Rhinol Laryngol. 1993;102:342– 348. 48. Carrol EN, Fine E, Ruff RL, et al. A four-drug pain regimen for head and neck cancers. Laryngoscope. 1994;104:694–700. 49. Keefe FJ, Manuel G, Brantley A, et al. Pain in the head and neck cancer patient changes over treatment. Head Neck Surg. 1986;8:169–176. 50. Gil Z, Smith DB, Marouani N, et al. Treatment of pain after head and neck surgeries: Control of acute pain after head and neck oncological surgeries. Otolaryngol Head Neck Surg. 2006;135:182– 188. 51. Bardiau FM, Taviaux NF, Albert A, et al. An intervention study to enhance postoperative pain management. Anesth Analg. 2003;96:179–185. 52. Gil Z, Cohen JT, Spektor S, et al. Anterior skull base surgery without prophylactic airway diversion procedures. Otolaryngol Head Neck Surg. 2003;128:681–685. 53. Shapiro A, Zohar E, Hoppenstein D, et al. A comparison of three techniques for acute postoperative pain control following major abdominal surgery. J Clin Anesth. 2003;15:345–350. 54. Massie MJ. Prevalence of depression in patients with cancer. J Natl Cancer Inst Monogr. 2004;2004(32):57–71. 55. Razavi D, Allilaire JF, Smith M, et al. The effect of fluoxetine on anxiety and depression symptoms in cancer patients. Acta Psychiatr Scand. 1996;94:205–210.
199
56. Fisch MJ, Loehrer PJ, Kristeller J, et al; Hoosier Oncology Group. Fluoxetine versus placebo in advanced cancer outpatients: A double-blinded trial of the Hoosier Oncology Group. J Clin Oncol. 2003;21:1937–1943. 57. Williams S, Dale J. The effectiveness of treatment for depression/depressive symptoms in adults with cancer: A systematic review. Br J Cancer. 2006;94:372–390. 58. Barsevick AM, Sweeney C, Haney E, et al. A systematic qualitative analysis of psychoeducational interventions for depression in patients with cancer. Oncol Nurs Forum. 2002;29:73–84. 59. Spiegel D, Bloom JR, Kraemer HC, et al. Effect of psychosocial treatment on survival of patients with metastatic breast cancer. Lancet. 1989;2:888–891. 60. Fawzy FI, Fawzy NW, Hyun CS, et al. Malignant melanoma. Effects of an early structured psychiatric intervention, coping, and affective state on recurrence and survival 6 years later. Arch Gen Psychiatry. 1993;50:681–689. 61. Richardson JL, Shelton DR, Krailo M, et al. The effect of compliance with treatment on survival among patients with hematologic malignancies. J Clin Oncol. 1990;8:356–364. 62. Bordeleau L, Szalai JP, Ennis M, et al. Quality of life in a randomized trial of group psychosocial support in metastatic breast cancer: Overall effects of the intervention and an exploration of missing data. J Clin Oncol. 2003;21:1944–1951. 63. Spiegel D, Spira J. Supportive-Expressive Group Therapy: A Treatment Manual of Psychosocial Intervention for Women with Recurrent Breast Cancer. Stanford, CA:Psychosocial Treatment Laboratory, Stanford University School of Medicine, 1991. 64. Winzelberg AJ, Classen C, Alpers GW, et al. Evaluation of an internet support group for women with primary breast cancer. Cancer. 2003;97:1164–1173.
14 Neurocognitive Assessment of Patients with Tumors of the Skull Base Mariana Witgert and Tracy Veramonti
INTRODUCTION
Pretreatment Studies Evidence of pretreatment cognitive impairment for at least some patients with skull base tumors has been documented. Meyers and colleagues (2000) (1) reported on a subset of nine patients with nasopharyngeal carcinoma who underwent neuropsychological evaluation prior to receiving paranasal sinus radiation. Although this small group, as a whole, performed better than patients who had previously undergone radiation, several patients did exhibit abnormal performance on objective measures of neurocognitive functioning prior to treatment. The authors noted that the presence of such pretreatment deficits may increase a patient’s vulnerability to adverse cognitive effects of radiation treatment. In a larger series of 28 patients with nasopharyngeal carcinoma who were evaluated prior to radiation therapy, neuropsychological test findings suggested impaired performance on measures of visuospatial reasoning (5). Additional evidence of pretreatment cognitive impairment was reported by Steinvorth et al. (2003) (6) in a study of 40 patients with skull base meningiomas. Pretreatment neuropsychological evaluation in these patients was notable for overall low average functioning, especially decreased information processing speed.
While neurocognitive symptoms are expected and often herald the diagnosis when patients present with intracerebral malignancies, patients with tumors of the skull base are also at inherent risk for development of neurocognitive symptoms, secondary to impingement and pressure effects on critical neuroanatomic structures as tumors in and around the cranial vault grow and crowd viable brain structures. Patients with anterior base of skull tumors (ABST) patients are further susceptible to cognitive symptoms as an unfortunate consequence of “incidental” neurologic involvement associated with common therapeutic treatments, for example, following radiotherapy to the paranasal sinuses (1). Furthermore, while surgical intervention has remained a “cornerstone” in the management of most patients with tumors of the skull base, more recently, improved patient outcomes have been associated with incorporation of other therapeutic strategies, such as chemotherapy in conjunction with radiotherapy (2). With the advent of more aggressive, multimodal treatment approaches, and improvements in radiation delivery techniques, the documentation of the cognitive sequelae of skull base cancer and its treatment remains essential.
Treatment Effects Even if the tumor itself does not result in cognitive impairments such as those described above, treatment may. Surgery may result in damage to normal tissue that surrounds the tumor, and yields more focal cognitive impairments that may resolve over time. In patients with skull base tumors, radiation is a more commonly utilized treatment than chemotherapy. Tolerance of radiation therapy varies among individuals. Important factors impacting individual tolerance include treatment factors such as dose, duration, and volume of brain irradiated, as well as patient variables such as age, genetic predisposition, preexisting neurologic diseases, systemic diseases that may predispose an individual to vascular injury, and concomitant chemotherapy (7). There is ample evidence that radiation therapy may be associated with the development of varying degrees of cognitive impairment both during and after treatment. Acute radiation encephalopathy develops within 2 weeks of initiation of treatment and is characterized by transient symptoms of headache, drowsiness, fever, and nausea (8). Early-delayed radiation encephalopathy develops 1–6 months after completion of radiotherapy and has been associated with transient declines in retrieval of verbal information in low-grade glioma patients (9,10). These impairments do not appear to be predictive of severe late-delayed, radiation-induced encephalopathy (9). The onset of late-delayed radiation encephalopathy can occur months to years afeter completion of radiation therapy and unlike acute and early-delayed effects,
WHY ARE ABST PATIENTS VULNERABLE TO NEUROCOGNITIVE DYSFUNCTION? Tumor Location The location of many skull base tumors places brain tissue, and consequently cognitive functioning, at risk of compromise secondary to compression of critical neuroanatomic structures in conjunction with tumor growth. Many skull base tumors are slow growing and become quite large before symptoms are noticed. For example, 62% of the 37 patients with olfactory groove meningiomas described by Turazzi et al. (1999) (3) and nearly 50% of patients with the same diagnosis in the Hentschel and DeMonte (2003) (4) study sample presented with tumors greater than 6 cm in diameter at the time of initial diagnosis. Symptom onset in these patients is commonly gradual until pressure effects lead to more significant problems. In the Hentschel and DeMonte (2003) (4) series of 13 patients, symptoms were often first noticed by family members and included a gradual onset of personality changes, such as apathy and akinesia, which is consistent with the subfrontal location of these tumors. Additionally, some patients experienced headaches, visual deficits (often inferior field cuts), and anosmia. Notably, anosmia is often reported only in hindsight, and is rarely a primary complaint of patients with olfactory groove meningiomas. 201
202
Witgert and Veramonti
tends to be irreversible (10). Impairments are associated with white matter changes secondary to demyelination and/or vascular damage, and occur on a continuum, ranging from mild to moderate cognitive impairments in attention and short-term memory to severe, progressive cognitive declines and dementia. In addition, endocrine dysfunction can be caused by damage to the hypothalamic–pituitary axis, leading to abnormalities in thyroid and human growth hormones (11), which have been associated with cognitive deficits (12– 14). While proton radiotherapy allows for relatively greater sparing of nontarget structures, and therefore may yield a reduction in radiation-induced morbidity, evidence suggests that radiation-induced endocrinopathy remains common even with advances in treatment modalities (15).
Posttreatment Studies A summary of studies investigating the effects of radiation treatment in skull base tumor patients is presented in Table 1. A review of Table 1 is notable for documented deficits in verbal and nonverbal memory, speed of mental processing, visual motor speed, executive functioning, fine motor coordination, attention and concentration, and visuospatial processing following radiation therapy to the skull base. In a study conducted by Steinvorth et al. (2003) (6), 40 patients with intracranial meningioma underwent neuropsychological evaluation prior to radiation treatment, following the first fraction of radiation, upon completion of treatment, and at 6 weeks, 6 months, and 12 months posttreatment. Examination of serial results from the follow-up evaluations revealed acute declines in verbal memory, regardless of the side of the brain that received radiation, with subsequent improvements over time. Follow-up in this study was limited Table 1
to 1 year and the authors stressed that the long-term effects of radiation treatment may not emerge until after this time period. The effects of radiation have also been investigated in other skull base tumor patients. Hua et al. (1998) (5) reported on the results of 27 patients with nasopharyngeal carcinoma who had undergone radiation to the brain. As compared to 35 healthy controls, patients who were evaluated postradiation performed worse on measures of auditory attention and concentration, verbal and nonverbal learning and memory, visuospatial reasoning, and manual dexterity. As compared to 27 patients with the same diagnosis who were awaiting radiation, the postradiation patients performed significantly worse on the measures of manual dexterity only. Additional information comes from Glosser et al. (1997) (16), who evaluated 17 patients with diagnoses of chordoma or low-grade chondrosarcoma before and after receiving radiation treatment. Following radiation, there was evidence of overall slowed performance, with a trend toward those patients who had received higher doses to be the slowest. There were no other noted declines. However, patients who were diagnosed with tumor recurrence or radiation necrosis were not included in the follow-up. The importance of long-term follow-up investigation of cognitive functioning was underscored by Meyers et al. (2000) (1), who performed neuropsychological assessment of 19 patients who had received therapeutic irradiation for skull base tumors at least 20 months prior. As a whole, patients performed below expectation on verbal memory, with 58% demonstrating difficulty learning new information and 69% demonstrating memory retrieval problems. Eighty percent of patients demonstrated accelerated forgetting of information
Summary of Studies Investigating Effects of Radiation on Cognitive Functioning in Patients with Skull Base Tumors
Study
Type of skull base tumor
n
Cognitive evaluation
Steinvorth et al., 2003
Intracranial meningioma
40
Glosser et al., 1997
Chordoma/low-grade chondrosarcoma
17
Pre-XRT; after 1st fraction; end of XRT, 6 weeks, 6 months, and 12 months post-XRT Postoperative/pre-XRT; post-XRT
Meyers et al., 2000
Paranasal sinus
28
n = 9, Pre- and post-XRT; n = 19, post-XRT only
Hua et al., 1998
NPC
90
n = 28, pre-XRT; n = 27, post-XRT; n = 35, healthy controls
Cheung et al., 2000
NPC
84
n = 53 NPC (31 w/TLN), post-XRT; n = 31 healthy controls
Cheung et al., 2003
NPC
50
≥1 year post-XRT
Lam et al., 2003
NPC
79
n = 60 NPC (40 w/TLN), ≥2 years post-XRT
a
Did not include patients with tumor recurrence or XRT necrosis at F/U. Abbreviations: NPC, nasopharyngeal carcinoma; XRT, radiation; TLN, temporal lobe necrosis.
Findings Acute decline in verbal memory, with subsequent improvements Post-XRT: slowed performance (trend = higher dose, slower performance); no other declinesa Impairments in learning and memory, visual–motor speed, and executive functioning Post-XRT vs. controls: worse auditory attention and concentration, verbal and visual learning and memory, visuospatial reasoning, and manual dexterity. Post-XRT vs. pre-XRT: worse manual dexterity only (trend for others). TLN associated with worse verbal and nonverbal memory, language, motor, and executive functioning, with relatively preserved intellectual, attention, and visual functioning. Trend for non-TLN patients to perform worse than healthy controls. Larger XRT-induced lesion volume associated with worse cognitive performance overall. Larger left-sided lesions associated with worse verbal memory and language. Larger right-sided lesions associated with worse visual memory. Both patient groups worse than controls on verbal memory.
Chapter 14: Neurocognitive Assessment of Patients with Tumors of the Skull Base
over time, a finding that was significantly associated with dose of radiation received; patients who received radiation doses greater than 60 Gy tended to have greater likelihood of impaired memory recall. Other areas of cognitive impairment included visual motor speed (35%), executive functioning (27–35%), and fine motor coordination (27–33%), with the overall pattern of findings consistent with frontal subcortical white matter dysfunction. The patients who had the worst outcome were those who received treatment prior to 1985 and who received higher doses of radiation. Studies investigating the late-delayed effects of radiation treatment have further confirmed the presence of cognitive deficits described by Meyers et al. (2000) (1), with those patients who develop radiation-induced necrosis demonstrating the weakest performance on measures of cognitive functioning. Cheung et al. (2000) (17) compared 53 patients with a diagnosis of nasopharyngeal carcinoma who had undergone radiation treatment with 31 healthy controls. Imaging evidence of temporal lobe necrosis (TLN) was observed in 31 of the 53 (58%) patients. TLN was associated with poorer verbal and nonverbal memory, as well as worse performance on measures of language, motor, and executive functioning. In addition, a nonstatistically significant trend was noted such that non-TLN patients generally performed worse than healthy controls but better than TLN patients. No significant differences between patients with and without TLN and controls were observed on measures of intellectual, attention, or visual functioning. In a subsequent study by Cheung and colleagues (2003) (18), 50 patients with nasopharyngeal carcinoma were evaluated at least one year after undergoing radiation therapy. Those with larger radiation-induced lesions were more likely to exhibit cognitive dysfunction. Left-sided lesions were associated with impairments in verbal memory and language, while right-sided lesions were associated with impairments in visual memory. Lam and colleagues (2003) (19) also reported impairments in verbal memory in a series of 60 patients with nasopharyngeal carcinoma (40 of whom had evidence of TLN) who were evaluated at least 2 years following completion of radiation. It is encouraging to observe that total dose of radiation treatment was the only major determinant of cognitive impairment in the Meyers et al. (2000) (1) study, whereas irradiated brain volume, laterality of radiation boost, and pretreatment chemotherapy were not significant factors. Current practice utilizes lower doses to the tumor as well as attempts to reduce exposure of the surrounding brain tissue. It is anticipated that new treatments will further improve dose density and limit incidental brain irradiation, and that cognitive complications will thus be minimized. However, as the effects of radiation treatment may not be evident for several years posttreatment, careful monitoring of cognitive function in patients over time remains necessary.
Frontal–Subcortical Circuits The pattern of neuropsychological findings associated with skull base tumors—which often includes impairments in processing speed, memory, visual motor speed, visuospatial reasoning, executive functioning, and fine motor coordination— is not surprising given the potential for disruption of frontal–subcortical circuits. In fact, ABST patients may be at particular risk for disruption of frontal–subcortical circuitry as tumors within the base of the cranial vault grow and disrupt these critical neuroanatomic networks (Fig. 1). In addition, previous research has documented that in patients with both cerebral and extracerebral malignancies, antineoplastic
203
treatments, including chemotherapies and/or radiotherapies are associated with a pattern of cognitive deficits consistent with frontal–subcortical white matter dysfunction, including neuropsychological deficits in executive functions, verbal retrieval, cognitive processing speed, and speeded motor coordination, as well as inefficiencies in learning and retrieval of stored information in the context of relatively well preserved memory consolidation processes (20,21). A thorough discussion of frontal–subcortical circuits is beyond the scope of this chapter; readers are referred to Mega and Cummings (1994) (22) for a review. Briefly, frontal–subcortical circuits mediate motor functioning, executive functioning, personality, and motivation. Executive functioning, regulated in part by the dorsolateral prefrontal circuit, is defined as “the ability to organize a behavioral response to solve a complex problem” and may involve tasks such as learning new information, organization and planning, shifting set, and inhibiting inappropriate responses. Changes in mood, in the form of depression and anxiety, have also been observed in association with damage to this circuit. Personality functioning is mediated by the lateral orbitofrontal and anterior cingulate circuits. The lateral orbitofrontal circuit is responsible for mediating socially appropriate behavior. Dysfunction in this circuit is most often manifested in the form of behavioral disinhibition or lack of empathy. In contrast, the hallmark of damage to the anterior cingulate circuit is apathy; patients may exhibit limited initiation of speech and movement and fail to display emotion.
NEUROPSYCHOLOGICAL ASSESSMENT To determine the pattern of a patient’s cognitive strengths and weaknesses prior to and following treatment, objective assessment of cognitive functioning via comprehensive neuropsychological assessment is necessary. Knowledge regarding the presence and pattern of neurocognitive impairments prior to treatment is essential in determining whether a specific anticancer therapy is associated with risks or benefits to cognitive functioning. Armed with baseline information, one can determine whether or not an individual has experienced significant change following treatment. In the absence of that information, patients can easily be misclassified as cognitively intact, when in fact they have declined significantly from premorbid levels of functioning. The reverse is also true; patients with a lower level of premorbid functioning may be incorrectly classified as having declined. The identification of changes in cognitive functioning helps inform differential diagnoses and serves as a starting point for development of pharmacologic and behavioral interventions to improve functional well-being. Wefel et al. (2004) (21) describe the complexities involved in neuropsychological assessment of patients with cancer, noting that while the administration of objective measures is relatively simple, the selection and interpretation of appropriate measures requires greater knowledge and skill. Test selection varies as a function of the question being asked; in the case of cancer patients in general and skull base tumor patients in particular, it is important to select measures that are sensitive to subtle changes in functioning. Measures should be reliable and valid, as well as robust to practice effects, as patients are often tested multiple times within a relatively short time span. Alternative test forms, when available, should be used. In addition to the above considerations, a thorough neurocognitive assessment includes an assessment of the indirect effects of cancer on physical and emotional well-being, and the consequent impact of such symptoms on cognitive
204
Witgert and Veramonti
Figure 1 Origins of three frontal–subcortical circuits. (Top left) Dorsolateral prefrontal circuit (blue). (Bottom left) Anterior cingulate circuit (red). (Right) Lateral orbitofrontal circuit (green). Source: Reprinted with permission from Mega & Cummings (2004).
functioning. Mental and physical fatigue can negatively impact cognitive functioning, as can numerous medications and associated medical complications. The reverse can also occur, with cognitive impairment leading to affective distress and fatigue (see Valentine & Meyers, 2001, for a more thorough review) (23). Cognitive effects of affective distress may vary from very mild impairments on effortful tests to significant deficits in visuospatial skills, executive functions, and psychomotor speed (24). It is important to note that in cancer populations, self report of cognitive impairment has been shown to correlate more intensely with fatigue and mood disturbance than with objective evidence of cognitive dysfunction, as assessed by standardized neuropsychological tests (25–28). Thus, a thorough neuropsychological assessment is needed to elucidate whether perceived difficulties are secondary to cancer-related cognitive dysfunction and/or affective distress. Of note, too often the assessment of cognitive functioning in clinical and research settings has been limited to brief screening measures, such as the Mini Mental State Examination (29). The poor sensitivity of this tool for patients with cancer has been documented (30). In fact, the inclusion of even brief neuropsychological assessment in research trials has been shown to be effective in identifying the risk and benefits associated with various anticancer therapies (20). The utility of neuropsychological assessment in patients with cancer is underscored by evidence demonstrating that cognitive impairment, when documented via
formal neuropsychological testing, predicts survival better than clinical prognostic factors alone in patients with primary brain tumors, leptomeningeal disease, and parenchymal brain metastases (20,31–33). Further, Meyers and Hess (2003) (34) demonstrated that cognitive performance was a more sensitive predictor of time to tumor progression than MRI, as cognitive decline occurred an average of 6 weeks prior to radiographic failure in 80 patients with glioblastoma multiforme and anaplastic astrocytoma. Neuropsychological evaluations tend to be well tolerated by cancer patients, with good compliance rates documented (33). As mentioned above, patients with skull base tumors are particularly vulnerable to disruption of frontal– subcortical networks. Patients and their caregivers may describe problems with organization, multitasking, and distractibility, as well as perceived forgetfulness, behavioral disinhibition, or apathy. Such complaints warrant a referral for neuropsychological evaluation, which should include a thorough clinical interview to elucidate perceived cognitive difficulties and associated functional impairments, as well as to define any compensatory strategies that are being used. The evaluation should utilize measures that are specifically selected to tap the cognitive domains most intimately associated with frontal–subcortical systems. For example, test selection might include measures of executive functioning, including set shifting, planning and organization, and response inhibition. In addition, within the domain of memory, frontal–subcortical dysfunction might result in impaired
Chapter 14: Neurocognitive Assessment of Patients with Tumors of the Skull Base
learning efficiency and retrieval. Speed of processing and fine motor control may also be impacted by frontal–subcortical dysfunction, and therefore are also appropriate for inclusion in a neuropsychological battery. Assessment of these cognitive skills should not occur in isolation, but rather in the context of tests of a comprehensive neuropsychological evaluation. A thorough assessment, with all included measures meeting the criteria described above, allows for evaluation of patterns of performance that can aid in differential diagnosis and treatment planning. Interested readers are referred to Lezak et al. (2004) (35) for a more in-depth discussion of specific neuropsychological measures and assessment procedures.
MANAGEMENT OF COGNITIVE SYMPTOMS: PHARMACOLOGIC AND BEHAVIORAL INTERVENTIONS Chan et al. (2003) (36) investigated the effects of vitamin E in 29 patients with nasopharyngeal carcinoma who had imaging evidence of unilateral or bilateral temporal lobe necrosis. Neuropsychological evaluations were performed before the initiation of treatment with vitamin E and 1 year later. The 19 patients who were treated with vitamin E performed better than 10 nontreated controls on measures of verbal learning and memory, visual memory, and cognitive flexibility. The authors acknowledged several limitations to this study, including that the patients were not randomized or blinded to their treatment. Further research is necessary to investigate the long-term safety and side effects of treatment with vitamin E, as well as the impact on cognition when treatment is discontinued and the mechanism by which vitamin E may improve cognitive functioning. Treatment with methylphenidate has been found to be beneficial in combating cognitive symptoms associated with treatment-related frontal–subcortical dysfunction in patients with primary brain tumors (37), such that patients demonstrated significant improvements in memory, psychomotor speed, visual–motor function, executive function, and fine motor speed. Other pharmacologic interventions that have been utilized in the neuro-oncology populations include donepezil to combat difficulties with attention and memory (38) and modafinil to alleviate fatigue and improve quality of life (39). Further research is needed to determine the effectiveness of these pharmacologic interventions for patients with ABST. Finally, although there is a dearth of evidence from prospective, randomized, controlled trials supporting their effectiveness, there are numerous anecdotal reports of goalfocused compensatory interventions and behavioral strategies being useful in minimizing the impact of cognitive deficits on daily life in patients with cancer. In fact, traditional rehabilitation disciplines (treating survivors of traumatic brain injury, stroke, or patients with dementia) have contributed a wealth of knowledge regarding evidence-based compensatory strategies that may be applicable to patients with cancer-related cognitive dysfunction (see Cicerone et al. 2000, 2005) (40,41). For example, external memory aids such as memory notebooks, user-programmable paging systems, and medication reminder systems have all been used to assist neurologically impaired patients compensate for difficulties with forgetfulness.
SUMMARY Patients with skull base tumors may be vulnerable to cognitive impairments secondary to disruption of frontal–
205
subcortical networks as a consequence of tumor growth and as an unfortunate side effect of antineoplastic treatments. Thorough neuropsychological evaluation is useful for assessing baseline (i.e., pretreatment) cognitive functioning, for monitoring the effects of cancer and cancer treatment on cognition, and for facilitating development and planning of behavioral and pharmacologic interventions to minimize the impact of symptoms on functional well-being and quality of life.
ACKNOWLEDGMENT We gratefully acknowledge Christina Meyers, Ph.D., ABPP, for her editorial suggestions. REFERENCES 1. Meyers CA, Geara F, Wong P-F, et al. Neurocognitive effects of therapeutic irradiation for base of skull tumors. Int J Radiation Oncology Biol Phys. 2000;46(1):51–55. 2. Prabhu SS, Demonte F. Treatment of skull base tumors. Curr Opin Oncol. 2003;15:209–212. 3. Turazzi S, Cristofori L, Gambin R, et al. The pterional approach for the microsurgical removal of olfactory groove meningiomas. Neurosurgery. 1999;45(4):821–825: discussion 825–826. 4. Hentschel SJ, Demonte F. Olfactory groove meningiomas. Neurosurg Focus. 2003;14(6):e4. 5. Hua M-S, Chen S-T, Tang L-M, et al. Neuropsychological function in patients with nasopharyngeal carcinoma after radiotherapy. J Clin Exp Neuropsychol. 1998;20(5):684–693. 6. Steinvorth S, Welzel G, Fuss M, et al. Neuropsychological outcome after fractionated stereotactic radiotherapy (FSRT) for base of skull meningiomas: A prospective 1-year follow-up. Radiother Oncol. 2003;69:177–182. 7. Taphoorn JB, Klein M. Cognitive deficits in adult patients with brain tumors. Lancet. 2004;3:159–168. 8. Keime-Guibert F, Napolitano M, Delattre J. Neurological complications of radiotherapy and chemotherapy. J Neurol. 1998;245:695–708. 9. Armstrong CL, Corn BW, Ruffer JE, et al. Radiotherapeutic effects on brain function: Double dissociation of memory systems. Neuropsychiatry Neuropsychol Behav Neurol. 2000;13:101–111. 10. Armstrong CL, Gyato K, Awadalla AW, et al. A critical review of the clinical effects of therapeutic irradiation damage to the brain: The roots of controversy. Neuropsychol Rev. 2004;14(1):65–86. 11. Darzy KH, Shalet SM. Hypopituitarism as a consequence of brain tumors and radiotherapy. Pituitary. 2005;8:203–211. 12. Dugbartey AT. Neurocognitive aspects of hypothyroidism. Arch Int Med. 1998;13:1413–1418. 13. Yudiarto FL, Muliadi L, Moeljanto D, et al. Neuropsychological findings in hyperthyroid patients. Acta Med Indones. 2006;38(1):6–10. 14. Deijen JB, Boer H, van der Veen EA. Cognitive changes during growth hormone replacement in adult men. Psychoneuroendocrinology. 1998;23:45–55. 15. Pai HH, Thornton A, Katznelson L, et al. Hypothalamic/pituitary function following high-dose conformal radiotherapy to the base of skull: Demonstration of a dose-effect relationship using dosevolume histogram analysis. Int J Radiation Oncology Biol Phys. 2001;49(4):1079–1092. 16. Glosser G, McManus P, Munzenrider J, et al. Neuropsychological functioning in adults after high dose fractionated radiation therapy of skull base tumors. Int J Radiation Oncology Biol Phys. 1997;38(2):231–239. 17. Cheung M, Chan AS, Law SC, et al. Cognitive function of patients with nasopharyngeal carcinoma with and without temporal lobe radionecrosis. Arch Neurol. 2000;57(9):1347–1352. 18. Cheung M, Chan AS, Law SC, et al. Impact of radionecrosis on cognitive dysfunction in patients after radiotherapy for nasopharyngeal carcinoma. Cancer. 2003;97(8):2019–2026.
206
Witgert and Veramonti
19. Lam LC, Leung SF, Chan YL. Progress of memory function after radiation therapy in patients with nasopharyngeal carcinoma. J Neuropsychiatry Clin Neurosci. 2003;15(1):90–97. 20. Meyers CA, Brown PD. Role and relevance of neurocognitive assessment in clinical trials of patients with CNS tumors. J Clin Oncol. 2006;8:1305–1309. 21. Wefel JS, Kayl AE, Meyers CA. Neuropsychological dysfunction associated with cancer and cancer therapies: A conceptual review of an emerging target. Br J Cancer. 2004;90:1691–1696. 22. Mega MS, Cummings JL. Frontal-subcortical circuits and neuropsychiatric disorders. J Neuropsychiatry Clin Neurosci. 1994;6:358–370. 23. Valentine AD, Meyers CA. Cognitive and mood disturbance as causes and symptoms of fatigue in cancer patients. Cancer. 2001;92(6):1694–1698. 24. Cull A, Hay C, Love SB, et al. What do cancer patients mean when they complain of memory problems? Br J Cancer. 1996;74:1674– 1679. 25. Castellon SA, Ganz PA, Bower JE, et al. Neurocognitive performance in breast cancer survivors exposed to adjuvant chemotherapy and tamoxifen. J Clin Exp Neuropsychol. 2004;26:955–969. 26. Jenkins V, Shilling V, Deutsch G, et al. A 3-year prospective study of the effects of adjuvant treatments on cognition in women with early stage breast cancer. Br J Cancer. 2006;94:828–834. 27. Schagen SB, Muller MJ, Boogerd W, et al. Cognitive dysfunction and chemotherapy: Neuropsychological findings in perspective. Clin Breast Cancer Suppl. 2002;3:S100–S108. 28. Houston WS, Bondi MW. Potentially reversible cognitive symptoms in older adults. In: Attix KD, Welsh-Bomer K, eds. Geriatric Neuropsychology Assessment and Intervention. New York, NY: Guilford Press, 2006:103–131. 29. Folstein MF, Folstein SE, McHugh PR. Mini-mental state. A practical method for grading the cognitive state of patients for the clinican. J Psych Res. 1975;12(3):189–198. 30. Meyers CA, Wefel JS. The use of the Mini-Mental State Examination to assess cognitive functioning in cancer trials: No ifs, ands, buts, or sensitivity. J Clin Oncol. 2003;19(1):3557–3558.
31. Sherman AM, Jaeckle K, Meyers CA. Pretreatment cognitive performance predicts survival in patients with leptomeningeal disease. Cancer. 2002;95(6):1311–1316. 32. Mehta MP, Shapiro WR, Glantz MJ, et al. Lead-in phase to randomized trial of motexafin gadolinium and whole-brain radiation for patients with brain metastases: Centralized assessment of magnetic resonance imaging, neurocognitive, and neurologic end points. J Clin Oncol. 2002;16:3445–3453. 33. Meyers CA, Smith JA, Bezjak A, et al. Neurocognitive function and progression in patients with brain metastases treated with whole-brain radiation and motexafin gadolinium: Results of a randomized phase III trial. J Clin Oncol. 2004;22(1):157– 165. 34. Meyers CA, Hess KR. Multifaceted end points in brain tumor clinical trials: Cognitive deterioration precedes MRI progression. Neuro-Oncology. 2003;5:89–95. 35. Lezak MD, Howieson DB, Loring DW. Neuropsychological Assessment. 4th ed. New York, NY: Oxford University Press,2004. 36. Chan AS, Cheung M, Law SC, et al. Phase II study of alphatocopherol in improving the cognitive function of patients with temporal lobe radionecrosis. Cancer. 2003;100(2):398–404. 37. Meyers CA, Weitzner MA, Valentine AD, et al. Methylphenidate therapy improves cognition, mood, and function of brain tumor patients. J Clin Oncol. 1998;16(7):2522–2527. 38. Shaw EG, Rosdhal R, D’Agostino RB, et al. Phase II study of donepezil in irradiated brain tumor patients: Effect on cognitive function, mood, and quality of life. J Clin Oncol. 2006;24(9):1415– 1420. 39. Nasir S. Modafinil improves fatigue in primary brain tumor patients (Abstract). Soc Neuro Oncol. 2003;5:335. 40. Cicerone KD, Dahlberg C, Kalmar K, et al. Evidence-based cognitive rehabilitation: Recommendations for clinical practice. Arch Phys Med Rehabil. 2000;81(12):1596–1615. 41. Cicerone KD, Dahlberg C, Malec JF, et al. Evidence-based cognitive rehabilitation: updated review of the literature from 1998 through 2002. Arch Phys Med Rehabil. 2005;87(3):1681– 1692.
15 Cerebrovascular Management in Skull Base Tumors Sabareesh Kumar Natarajan, Basavaraj Ghodke, and Laligam N. Sekhar
venous infarction. The vein of Labb´e often consists of multiple veins. One may drain into a tentorial sinus, and another may drain far anteriorly into the transverse sigmoid junction. This will have an influence on the approach being planned.
INTRODUCTION Cranial base tumors often involve the basal arteries and veins. Successful management of these lesions requires a multidisciplinary approach to patient care. An important component to this approach is the management of the cerebral vasculature, which includes preoperative assessment of the cerebral vasculature, preoperative embolization, arterial and venous preservation, and reconstruction during resection and the management of vasospasm after resection.
Preoperative Embolization We attempt to embolize all cranial base meningiomas, whenever feasible, to reduce the blood loss and operative time. When embolization is adequately performed, tumor necrosis or softening may also occur making the operation easier. Limited blood loss makes surgical resection safer and improves visualization. Necrosis often leads to softening of the tumor, which not only facilitates resection but also reduces the forces transmitted to adjacent sensitive neural structures. The arteries commonly embolized include meningohypophyseal branch of the ICA; branches of the external carotid artery—the sphenopalatine artery, middle meningeal artery, accessory meningeal artery, internal maxillary artery, ascending pharyngeal artery, and other; and rarely meningeal branches of the vertebral artery (VA). Tumor embolization is performed under local (ECA branches) or general (ICA branches) anesthesia. To facilitate the selective catheterization of the small branches, a R Renegade (Boston Scientific, Natick, MA) (for larger pediR cles such as ascending pharyngeal artery) or Marathon (ev3, Irvine, CA) (for smaller pedicles such as meningohypophyseal trunk) microcatheter is used. A manual injection of undiluted, nonionic contrast agent is performed with a 1.0 mL syringe, starting very slowly until the contrast agent becomes visible and then the rate of injection is increased slowly. If tumor vascularity is apparent, the injection is repeated, increasing the rate of injection to possibly identify reflux into the carotid siphon. Filming is biplane to evaluate potential cross filling of contralateral branches. An idea of the force required to cause reflux into cerebral arteries is thus obtained prior to the embolization procedure. Embolization is ideally performed 3 to 7 days before the planned procedure. The goal of embolization is to permeate the interstices of the tumor with particles, and occluding the feeding arteries at the very end, if they are large. Embolization material consists of particles of polyvinyl acetyl foam (PVA, Cook, Inc, Bloomington, IN) suspended in undiluted, nonionic contrast agent injected slowly. The size of the particle used depends on the potential for reflux and supply to cranial nerves from the pedicle being embolized. The average size of the particles used for embolization is 150 to 250 µ, with larger particles 250 to 400 µ being used for the embolization of the ascending pharyngeal artery to avoid occlusion of branches feeding cranial nerves X and XI. Additionally, small gelfoam pledgets are used to block arteries at the end of the procedure. In hypervascular tumors, liquid embolic
Preoperative Imaging The MRI scan is an important study. In short T2-weighted images, the presence of arterial encasement can be seen clearly. The extent of encasement and any narrowing of the artery is noted. Cerebral angiography is a must in all patients with arterial encasement. All major vessels must be studied in order to evaluate potential collateral sources, along with the evaluation of venous system. In case of internal carotid artery (ICA) encasement, we perform angiography with ipsilateral CCA compression. In a patient in whom surgical occlusion is planned, a carotid compression arteriogram with contralateral carotid and vertebral injection is performed in order to evaluate collateral flow through the anterior communicating artery (ACOM) or the posterior communicating artery. This information is used to decide whether the ECA (no cross flow) or ICA (cross flow present) should be used as the donor artery, which gives the surgeon some idea of collateral sources, to assess tolerance to temporary occlusion during a bypass procedure. Before operating on basal tumors or tumors located near the torcula, transverse and sigmoid sinuses, vein of Labb´e, or the straight sinus, angiography with filming of the venous phase is important to provide information about the adjacent veins and collaterals. The size and dominance of the transverse or sigmoid sinus and collateralization through the torcula can alter the surgical approach by revealing a very large sigmoid sinus or high riding jugular bulb, either of which can significantly affect the amount of exposure. Similarly, a small sigmoid sinus with excellent collaterals provides a wider presigmoid exposure with less consequences should sigmoid sinus occlusion occur. This also helps in decision making regarding possible venous occlusion. Configuration, relative sizes, and anastomotic relationships of the veins draining the temporal lobe are extremely important. In most patients, the veins of Labb´e and the superficial middle cerebral veins have a seesaw relationship, commonly being relatively equal in size. In some patients, one or more of these veins will be larger in size. The consequences of occluding a large or dominant vein must be factored into the surgical planning to protect them and avoid 207
208
Natarajan et al.
The use of bypasses for skull base tumors has declined in frequency due to the increasing use of radiosurgery for tumor remnants left around the carotid artery during surgery for basal meningiomas. However, this technique remains a valuable tool in the management of skull base tumors such as recurrent meningiomas, recurrent chordomas, chondrosarcomas, and other malignant tumors (1–7).
Operative Technique in a Tumor with Subarachnoid Encasement
Figure 1 Illustration showing cervical ECA to MCA-M2 bypass.
R agents such as Onyx , n-BCA, or coils are used for flow reduction. The risks of embolization such as skin necrosis, cranial nerve dysfunction, stroke, blindness should be weighed against the perioperative and postoperative advantages. Embolization of some large or giant tumors may result in tumor swelling, which may precipitate emergent surgery.
MANAGEMENT OF ARTERIAL ENCASEMENT Intracranial skull base tumors frequently encase the basal arteries, ICA, VA, basilar artery (BA), and their branches. When the encasement is in the subarachnoid segment of the artery, the tumor can be dissected away from the artery due to existence of an arachnoid plane around the tumor. Such a plane is absent in cases of prior surgery or radiotherapy and such dissection is not feasible. When the tumor involves the extradural segment, dissectability depends upon the pathology of the tumor and the oncologic goals of the surgery. Schwannomas or cavernous hemangiomas are always dissectible. Meningiomas can often be dissected when the vessel is only encased, but usually not when the vessel is encased and narrowed. In patients with malignant tumors such as adenoid cystic carcinoma, the ICA has to be resected because of tumor invasion.
Figure 2
Illustration showing cervical ECA to supraclinoid ICA bypass.
Proximal control of the exposed artery is obtained, followed by adequate exposure of the tumor without brain retraction, and then distal exposure of the artery beyond the encasement. The artery is then traced through the tumor from both sides, with frequent debulking. Suction, bipolar cautery, fine dissectors, and microscissors are used for dissection of the artery. In case of an artery that has multiple perforators, the surface without perforators is dissected first, followed by the portion containing perforating branches (see Case Studies).
When to Bypass After Vascular Sacrifice? When the intracranial ICA or VA is occluded, whether or not a bypass is needed in all patients is controversial. A selective approach bases this decision on a balloon occlusion test, with monitoring of cerebral blood flow by SPECT, TCD, or angiography. A universal approach recommends a bypass in all patients. Based on a review of patients of senior author who were not revascularized and suffered strokes and the reports of other surgeons with a similar experience (8–11), a universal approach is presently followed if the ICA has to be occluded for tumor cases. In regards to the VA, if a vessel is markedly non dominant, it needs not be reconstructed. However, an equidominant or dominant VA must be reconstructed. If the BA is damaged, it must always be reconstructed, although when the patient has two large posterior communicating arteries, the patient may not suffer a stroke because of good collaterals. In the event of unexpected intraoperative injury to major arteries, it is best to reconstruct the vessel using either a local, regional, or extra-intracranial bypass technique, since the adequacy of collateral circulation cannot be determined.
Figure 3 (A, B) Terminal stitches are placed at the heel and the diametrically opposite ends. Also note the fish-mouthing of the STA. (C) One side is anastomosed with continuous sutures and (D) the opposite side with interrupted sutures.
Chapter 15: Cerebrovascular Management in Skull Base Tumors
130 79 7 7 5 32
raises the patients’ BP by about 20% during temporary occlusion, and places the patient in burst suppression with propofol, to protect the brain. Approximately 4000 to 5000 units of heparin are also administered during the vascular occlusion. Epidural hemostasis must be excellent, to prevent excessive oozing after heparinization. The heparin is not reversed at the end of the procedure.
29 101
Bypass Techniques High-Flow Bypass
Table 1 Revascularization for Tumors 1988–2006 Number of patients Meningiomas Chondrosarcoma Chordoma Adenoid cystic carcinoma Miscellaneous Type of graft Radial artery graft Saphenous vein graft Extent of resection Gross total resection Incomplete resection Graft patency Immediate patency rate Delayed graft occlusion Mortality Due to Surgery Major stroke despite patent graft Preoperative deficits and morbidity after surgery Due to disease progression/recurrence
209
82 (63%) 48 (37%) 124/130 (95.4%) 2 (managed by revision, patent at follow-up)
RAG or SVG is used for bypass, based on preoperative Duplex imaging. For a tumor involving the ICA, ECA or cervical ICA to MCA-M2 segment bypass is preferred (Fig. 1). When the MCA vessels are small, then the supraclinoid ICA may be used as a recipient vessel (Fig. 2). The cervical ICA is
2/130 (1.5%) 1 1 17/130 (13.1%)
Choice of Bypass Grafts Bypasses may be divided into two groups as: replacement bypasses (e.g., radial artery graft/saphenous vein to replace the ICA/VA) and augmentation bypasses (e.g., STA-MCA bypass performed in a patient with brain ischemia secondary to ICA occlusion). Most of the bypasses performed for skull base tumors are replacement bypasses with radial artery, saphenous vein, or rarely, the superficial temporal artery. A local repair of an injured artery may also be performed in some cases. The radial artery provides flow rates between 50 and 150 mL/min acutely and the flow can increase significantly over the ensuing days as measured by Duplex ultrasound (3). The radial artery is easier to harvest than the saphenous vein. However, a major problem with its use is the occurrence of postoperative spasm, which can be prevented by the use of the pressure distension technique (1,3,12). The saphenous vein graft is used when the radial artery is not available as a suitable vessel. It may also be used in cases where the radial artery graft, for one reason or the other, has failed. In children below the age of 12, where the radial artery graft is small in diameter, the saphenous vein graft is the only alternative. Graft flow in the saphenous vein has been measured from 100 to 250 mL/min. Because of the high flow through the saphenous vein graft, there may be a flow mismatch when it is anastomosed into the middle cerebral artery or the posterior cerebral artery, which could lead to turbulence and graft flow problems. It is more prone to kinking the distal anastomotic site and is technically more difficult to perform because of its thicker wall.
(A)
Anesthesia, Monitoring, and Preparation If a bypass procedure is a strong possibility, preoperative Duplex imaging of the radial arteries and saphenous veins, and an Allen test of the hands are performed to choose the vessel that may be used for bypass. The patient receives aspirin 82 mg preoperatively. Intraoperative monitoring of electroencephalogram, somatosensory evoked potentials, and motor evoked potentials is employed. Total intravenous anesthesia is used to allow monitoring of motor evoked potentials. If a bypass is performed, the anesthesiologist
(B)
Figure 4 (A) Normal type of vein of Labb´e. (B) Very anteriorly draining vein of Labb´e into a tentorial sinus. (continued on page 208).
210
Natarajan et al.
(C)
(D)
Figure 4
(C) Variation of anatomy of vein of Labb´e seen in angiogram. (D) Postoperative venous infarction of the temporal lobe.
exposed in the neck. The tumor is exposed after a craniotomy and an orbital or orbitozygomatic osteotomy. After tumor inspection, and in some cases, after attempted removal of the tumor, a decision is made to proceed with the bypass. Accordingly, the radial artery (the entire artery from the brachial artery bifurcation to the anterior wrist) or the saphenous vein (in the upper leg and lower thigh) is removed, flushed with heparinized saline, and distended under pressure to relieve
vasospasm. The distal anastomosis is performed first, to the MCA (M1 bifurcation or M2 segment) or to the supraclinoid ICA. This is followed by the proximal anastomosis to the ECA (if collateral circulation is poor) or to the ICA (if some collaterals are present). If the flow through the grafts is satisfactory by Doppler/intraoperative angiography, then the ICA is trapped between clips. The operation is stopped at this stage. A postoperative angiogram is performed the next day.
Chapter 15: Cerebrovascular Management in Skull Base Tumors
211
(A)
Figure 5 Technique of inserting a butterfly needle into the sinus following placement of temporary clip to occlude the sinus.
For VA replacement, a RAG or saphenous vein interposition graft is employed. An extreme lateral retrocondylar or partial transcondylar approach is used. Proximal anastomosis to the VA is done at the level of C1–C2. If the distal anastomosis is distal to the PICA, then the PICA may be reimplanted or a PICA to PICA anastomosis performed. Alternatively, the PICA may be occluded if there is good collateral flow from the distal vessel. For BA injury, VA or ECA to PCA-P2 segment bypass is performed with RAG or saphenous vein
(B)
(C)
Figure 7 Repair of a superior sagittal sinus by graft. (A) Tumor invading the sinus. (B) Part of the patch graft is sutured to the sinus wall with tumor in situ. (C) Patch graft sutured after tumor excision.
graft. A temporal craniotomy with a zygomatic osteotomy or a petrosal approach is used for the exposure of the PCA.
STA to MCA Bypass
(A)
(B)
Figure 6 (A, B) Direct repair of the sigmoid sinus after resection of a meningioma, which had encased the sinus.
The STA is exposed by a direct cut-down technique. The course of the vessel is traced by Doppler ultrasound or frameless neuronavigation and marked on the scalp. Working under the operating microscope, dissection is started distally, and the vessel is traced proximally. A small cuff of connective tissue is left around the artery. The vessel is left in situ until the bypass procedure. A T-incision is created from the skin incision made to expose the artery to facilitate the muscle dissection and the performance of a small pterional craniotomy. A middle cerebral branch in the distal Sylvian fissure (M3 branch), the largest temporal or parietal cortical branch relatively free of perforators, is used for anastomosis. Ideally, the recipient vessel has to be at least 1.5 mm in diameter but can be as narrow as 1.0 mm. The recipient vessel is dissected free of its arachnoidal covering, and a small rubber dam is placed under the artery. The STA is divided and an
212
Natarajan et al. Table 2
Indications for Reconstruction of Cerebral Venous Sinuses
Status
Decision
Collaterals Excellent collaterals Marginal collaterals Poor or no collaterals Sinus occlusion One wall involved, sinus occlusion <50% Two walls involved, sinus occlusion >50% Three walls involved, sinus occlusion >90%
Reconstruction unnecessary; provides practice for surgeons Reconstruction recommended Occlusion dangerous; reconstruction if there is accidental injury Resection, resuture or a small patch Resection possible, vein patch for repair, preserve collaterals Test occlusion with pressure monitoring, sinus repair if pressure > 5 mmHg
oblique arteriotomy with slight fish-mouthing of the STA is done. Anastomosis to the MCA is done using diametrically opposing sutures at the ends, and interrupted 9/0 or 10/0 nylon sutures (Fig. 3). Prior to tying the last suture, the lumen is flushed with heparinized saline and the suture tightened. Flow through the STA is checked with a Doppler probe.
Staged Operations When a bypass is performed, staged operations are usually preferred. Craniotomy, exposure osteotomies, and bypass are done during the first surgery, and is followed by tumor resection in the second operation 3 to 7 days later. This is because both aspirin and intravenous heparin are administered during the first operation, and tumor resection and repair of skull base may need extensive work on the same day. Exceptions to this are removal of tumors around the VA, which may require a short segmental bypass following removal of a small tumor, with critical vascular encasement.
to follow the volume flow through the graft. The systolic BP should be maintained below 140 mmHg for 2 to 3 days. Graft occlusion may occur, either during the operation or within the first 24 hours after surgery. If occlusion occurs during surgery, the problem is corrected. If the occlusion occurs postoperatively, the patient needs to be reoperated. Rarely, endovascular thrombolysis can be done. Vasospasm occurs occasionally with radial artery grafts (despite the intraoperative pressure distention technique) and can be successfully treated by endovascular angioplasty. Following discharge (7–10 days), the patients are kept on statins for life as well as on aspirin (81 mg PO once daily) for life in case of vein graft and at least for 1 year in case of radial artery grafts. The graft is followed by 3D CTA or MRA at 3 months postoperatively and also by using graft flow measurements with Duplex imaging. Subsequently we recommend a 3D CTA at one year, and thereafter every 1 to 2 years.
Results Postoperative Management Postoperative monitoring of graft patency is usually done by Doppler evaluation. A CT angiogram is done immediately after the operation if an intra-arterial DSA was not performed to evaluate the graft. Otherwise, a postoperative DSA is performed at 12 hours after surgery. Patients are started on aspirin, 81 mg once daily and statins. Patients with saphenous vein grafts are maintained on subcutaneous heparin, 5000 U every eight hours for 3 days in addition to aspirin and statins. Duplex ultrasound studies are performed
One hundred and thirty patients underwent bypasses for tumors from 1988 to 2006 (79 for skull base meningiomas, 7 chondrosarcomas, 7 chordomas, 5 adenoid cystic carcinomas and other tumors like osteogenic sarcoma, schwannoma, hemangiomas, and hemangiopericytomas) (Table 1). The
(A)
(A)
(B)
Figure 8 (A) Tumor encased and invaded the sigmoid sinus to narrow the sinus. (B) Part of the sinus was resected along with the tumor, and an interposition graft (saphenous vein) was placed (end-to-end anastomoses).
(B)
Figure 9 (A) Tumor invading sinus. (B) Tumor resected with sinus and interposition graft (radial artery) placed (end-to-side anastomoses).
Chapter 15: Cerebrovascular Management in Skull Base Tumors Table 3
213
Results of Venous Reconstruction from 1993 to 1999
Type of vein/venous sinus reconstructed Cortical vein Resuture graft SVG RAG Transverse sinus (pineal region) Division only Division + SVG Sigmoid sinus–jugular vein Division + resuture Repair after tumor excision SVG
No. of patients
No. of patients in which they are patent finally
1 1 1
1 1 1
5 1
NA 1
4 2 2
4 1 2
immediate patency rate for bypasses was 95.4%. Gross total resection was achieved in 82 (63%) of patients. There were 29 radial artery grafts and 101 saphenous vein grafts used. Two patients had delayed graft occlusions (after 2 years),
which were revised. Sixteen patients had disease progression or recurrence and died. One patient who was wheel chair bound with multiple lower cranial nerve palsies died 8 months after surgery. One patient had a major stroke despite functioning of the graft and died after 7 days.
PRESERVATION AND RECONSTRUCTION OF VEINS AND SINUSES Major cerebral veins are usually at risk during surgery for skull base tumors, because of displacement of veins from fixed drainage sites in the brain and/or their division to approach a lesion. When the venous outflow is compromised
Figure 10 Preoperative MRI images showing a giant petroclival Meckel’s cave and cavernous sinus meningiomas with severe brainstem compression displacing and adherent to the basilar artery.
Figure 11 Postoperative MRI images showing subtotal removal of the tumor with preservation of the basilar artery.
214
Natarajan et al.
Figure 13 Preoperative angiogram showing an isolated left anterior circulation. Figure 12 Preoperative MRI showing a recurrent meningioma involving orbit, cavernous sinus, and supracavernous regions with encasement and narrowing of the internal carotid artery.
due to a lack of adequate collateral circulation, venous infarction follows, with swelling, hemorrhage, and neuronal death. The clinical consequences, which can often be disastrous, will depend upon the region of involvement of the brain and size of the venous structure occluded. The consequences of cerebral venous sinus occlusion also depend upon the availability of collateral circulation. When such collaterals are not available, papilledema and visual loss and a pseudotumor cerebri syndrome are observed in milder cases, whereas severe diffuse brain swelling, coma, and death may be observed in severe cases. Acute venous or venous sinus occlusion is potentially very dangerous, whereas slow and chronic venous or venous sinus occlusion is better tolerated. Even in such patients, some neurologic manifestations may follow when the collaterals are poor. Veins have thinner walls and are not as tortuous as arteries, giving them less room for manipulation before rupture. Basal operations stretch the veins and put them at risk for rupture. To avoid rupture of a cerebral vein it should be stretched minimally either by minimizing the extent of brain retraction or by releasing it from adhesions to allow its lengthening. With basal lesions, veins at greatest risk for rupture are the temporal tip draining veins, and the vein(s) of Labb´e. In these instances, the surgeon must be aware of any aberrant venous anatomy before surgery to avoid major problems. In the majority of patients, the temporal tip draining veins can be divided without adverse consequences. However, when the Sylvian vein is very large or if the vein of Labb´e is absent due to prior surgery or is very small because of an anatomic variation, then the superficial middle cerebral (Sylvian) may not be safely occluded. In such a situation, a temporary clip can be placed on the concerned vein, and the brain is
(A)
(B)
Figure 14 Illustration showing radial artery graft bypass from the ECA to the MCA-M2 and a saphenous vein graft from the cervical ICA to a different branch of MCA-M2 segment.
Chapter 15: Cerebrovascular Management in Skull Base Tumors
215
Figure 15 Postoperative angiogram showing radial artery graft bypass from the ECA to the MCA-M2 and a saphenous vein graft from the cervical ICA to a different and slightly larger branch of MCA-M2 segment.
observed for swelling for 5 to 10 minutes. If the vein cannot be occluded safely, in many patients a change in the surgical approach or a small corticectomy will allow the operation to be performed. The vein(s) of Labb´e [Fig. 4(A)–4(D)] are at risk for injury during subtemporal and transpetrosal approaches (13–18). The partial labyrinthectomy petrous apicectomy transpetrosal approach, the translabyrinthine approach, and the total petrosectomy approach all move the surgeon anteriorly from the drainage point of the vein of Labb´e. However, in some patients, these strategies may not be enough to prevent excessive stretching of the vein. In some patients, the tentorium may be divided with minimal brain retraction, and then the retractor can be placed on the tentorium rather than the temporal lobe to prevent venous stretching. When the vein is very large and dominant with a very anterior drainage site [Fig. 4(B)], then the surgical approach may have to be changed to the retrosigmoid (or retrosigmoid + orbitozygomatic with frontotemporal craniotomy) to prevent venous injury, especially on the dominant side.
Figure 16 Postoperative MRI showing complete removal of the tumor.
brain. The only venous sinuses that may safely be occluded in most patients are the cavernous sinuses, the superior petrosal sinuses, and the nondominant, well-collateralized transverse and sigmoid sinuses. Occlusion of the cavernous sinus can usually be performed without adverse effects on vision and the orbit due to the presence of many collateral drainage channels from the orbit.
Venous Reconstruction The indications for venous reconstruction are accidental injury to large veins, if brain swelling is noted after the occlusion of a vein, or after the injury to any deep vein. In such patients, the easiest reconstruction may be by direct suture, using 8–0 nylon sutures. If the anastomosis is under tension, some of the tension can be released by dural mobilization. Direct repair is usually successful, even if the repaired vein is slightly stenotic. When a segment of the vein is missing and direct repair is difficult, a segment of saphenous vein from the leg, or a vein from the forearm or the neck, or the radial artery may be used as an interposition graft (17). Postoperative thrombosis is the main problem with venous reconstruction and may occur because of injury to the endothelium of the transplanted vein and the slow blood flow through the vein in general. To prevent this, we give the patients 4000 U of intravenous heparin during the reconstruction procedure, subcutaneous heparin during the first 7 postoperative days (5000 U q8h), and aspirin 325 mg daily thereafter for 2 to 3 months.
Cerebral Venous Sinuses Cerebral venous sinuses transmit a large volume of venous blood from the brain. The patency of the venous sinuses is very important to preserve the functional integrity of the
Intraoperative Sinus Occlusion Test During surgery, before occluding a sinus, a test occlusion must be performed (Fig. 5). To do this, the intrasinus pressure is measured by inserting a 20 gauge butterfly needle connected to a pressure transducer. The normal venous sinus pressure should be less than 15 mm Hg, depending upon the position of the head. After a stable reading is obtained, a temporary clip is applied on the venous sinus at the appropriate point of expected occlusion. Observation of the brain or cerebellum for swelling and the evoked potentials and intrasinus pressure is performed for at least 5 minutes. Intrasinus pressure is the most sensitive indicator of the three, but cerebellar swelling may occur very quickly. If brain swelling occurs, evoked potentials change, or intrasinus pressure increases by more than 5 mm Hg, then the temporary clip is removed and the sinus cannot be occluded. If the initial intrasinus pressure is above 15 mm Hg but there is no significant increase in the pressure, the sinus may be occluded, but continuous monitoring of the pressure must be done during the rest of the operation because a delayed increase in intrasinus pressure may occur and necessitate reconstruction. Preoperative occlusion tests of the venous sinuses are not safe because the clinical response is delayed, and the effects are not fully reversible.
216
Natarajan et al.
(A)
(B)
(C)
Figure 17 Preoperative MRI (A, B) showing chondrosarcoma involving the cavernous sinus and infiltrating the brain stem. (C) Illustration showing infiltration of cavernous ICA by the tumor.
Chapter 15: Cerebrovascular Management in Skull Base Tumors
217
Figure 18 Illustration showing radial artery bypass to the MCA and removal of the tumor after trapping the ICA.
Reconstruction of Venous Sinuses Direct Repair When a small portion of the circumference of a venous sinus is involved by a tumor, direct repair is recommended. In such patients, the tumor is excised, and the sinus is repaired with 5– 0 prolene [Fig. 6(A) and 6(B)] sutures either by direct suturing or with a patch of dura mater or saphenous vein. The graft is sutured onto some of the sinus wall before removal of the tumor [Fig. 7(A) and 7(B)]. After removal of the tumor, the sinus may be allowed to bleed if it is a small rent, occluded with finger pressure or temporary clips if some collaterals exist, or occluded with a balloon shunt if high flow exists through the sinus [Fig. 7(C)]. If the repair is likely to take more than 10 minutes, then the patient will need to be heparinized. When the sigmoid sinus is divided to improve the exposure of the tumor, direct repair may be performed with 6–0 prolene sutures.
Graft Reconstruction Graft reconstruction of the sinus is performed in cases of total segmental defect (17), which cannot be repaired directly. The indications for such sinus repair are shown in Table 2. When the sinus to be repaired is large (≥1 cm diameter), the saphenous vein extracted from the thigh is used [Fig. 8(A) and 8(B)]. When the sinus has been previously partially occluded by the tumor, the radial artery is used because it tends to stay open even while the flow rate is low. Because of the discrepancy in size, an end-to-side technique is used for radial artery grafts [Fig. 9(A) and 9(B)], whereas an end-to-end technique is used for saphenous vein grafts.
Our Approach for Tumors Involving Cerebral Veins and Venous Sinuses Results of cerebral venous and venous sinus reconstruction in the 6 years from 1993 to 1999 for meningiomas are shown in Table 3. The principles that we follow in managing patients
with tumors involving the cerebral veins and venous sinuses are as follows: 1. Meningiomas are treated only if they are symptomatic, when they are found to involve the venous sinuses. 2. There are two options—radical resection with the involved sinus or conservative resection and radiosurgery. The decision to go ahead with radical resection involves the basic principles of any tumor resection such as age, comorbidities, and regions involved, major vascular and cranial nerve involvement. Once a decision is made to do a radical resection with the sinus, the principles of whether reconstruction is required or not is followed as described above. 3. Peeling of the tumor from the sinus can be done if only one of the walls is involved and the wall is reconstructed as described above. 4. If two walls are involved, the tumor may be still removed and repaired with a dural or venous patch. 5. If three walls or more than 50% of the sinus is involved, then sinus reconstruction may be required. 6. If the sinus is completely occluded preoperatively, no repair is needed. However, the surgeon must be prepared for sinus repair, if some flow exists intraoperatively. 7. Radiosurgery may cause delayed sinus thrombosis with brain edema and seizures. All veins should be preserved.
Vasospasm After Cranial Base Tumor Resection Cerebral vasospasm is well known to occur after various cerebral neurosurgical events that cause subarachnoid hemorrhage. Cerebral vasospasm can also occur after cranial base tumor resection (19). Vasospasm manifests clinically 7 to 30 days postoperatively. Pituitary tumors and craniopharyngiomas have a higher incidence of vasospasm, but the reason for this is not clear. Delayed neurologic deterioration in a patient who has undergone cranial base tumor surgery not explained by an intracranial mass lesion should be promptly investigated with angiography. If vasospasm is diagnosed,
218
Natarajan et al.
(A)
(B)
Figure 19 Postoperative MRI showing complete removal of the tumor. A (Axial), B (sagital).
it should be treated aggressively with hypertensive, hypervolemic therapy and if necessary, endovascular angioplasty.
CASE STUDIES Case Study 1 Meningioma with Subarachnoid Vascular Encasement A 43-year-old woman (Fig. 10) with a giant petroclival Meckel’s cave and cavernous sinus meningioma with severe brainstem compression, underwent a subtotal resection (Simpson grade III) of the tumor. This patient had a history of transient diplopia and facial numbness as well as some numbness involving the right side of the body. Frontotemporal craniotomy and a retrolabyrinthine transpetrosal approach were used to approach the tumor. The superior cerebellar artery had two branches encased by the tumor, which were dissected free. There were two brainstem perforators densely
adherent and partially encased just inferior to the superior cerebellar artery. These required very tedious and difficult dissection. The tumor was completely removed from the clival dura. The posterior communicating artery was preserved. The BA was completely dissected, and the tumor was gradually removed piecemeal in a complete fashion in this area. Tumor was left behind in the region of the Meckel’s cave, around the sixth nerve and in the posterior part of the cavernous sinus (Fig. 11). This was treated with radiosurgery. She has no postoperative neurologic deficits and required a ventriculoperitoneal shunt for hydrocephalus.
Case Study 2 Cavernous Sinus Meningioma: Postradiation Bypass A 47-year-old woman had headaches and ptosis was diagnosed to have a left-sided cavernous sinus meningioma (Fig. 12). She had fractionated, stereotactic, conformal
Chapter 15: Cerebrovascular Management in Skull Base Tumors
(A)
219
(B)
Figure 20 (A) Axial and (B) sagittal contrast-enhanced T1 image of a magnetic resonance imaging scan showing the petroclival meningioma in Case 1.
(A)
(B)
Figure 21 (A, B) Venous phase of angiogram showing prominent vein of Labb´e and dominance of the right lateral sinus.
Figure 22 Diagrammatic representation of the vein graft technique. Fish-mouth opening is made in the vein as well as in the graft to reduce the amount of stenosis later on. Also, a part of the sinus wall was removed.
220
Natarajan et al.
Figure 23 Diagrammatic representation of the vein graft technique (continued). The anastomosis is completed, and a pad of fat supports the graft.
radiotherapy 5 years ago; since that time, the tumor had been growing steadily and she has had spasm of the jaw. On examination, the patient had no vision in her left eye, partial ptosis, near complete ophthalmoplegia and restricted opening of the mouth without any sensitivity of the TM joint. MRI and angiography revealed an intracavernous and supracavernous tumor encasing and narrowing left ICA. The left ICA had a small ACA, but the circulation was nearly isolated in the left hemisphere (Fig. 13). She underwent RAG bypass from the cervical ECA to the M2 segment of the MCA followed by
SVG bypass grafting from the ICA to a different and slightly larger branch of the M2 segment of the MCA (Fig. 14). She had a temporal lobe clot on the immediate postoperative CT, which was evacuated. After 3 days, she had a gross total resection of the tumor. She had a CSF leak postoperatively, which was repaired by endonasal approach. She had no new neurologic deficits. She had no tumor remnant by MRI (Fig. 15) and patent grafts on CT angiogram after 10 months (Fig. 16).
Case Study 3 Recurrent Chondrosarcoma: Emergent Bypass Figures 17–19 show a patient with a recurrent chondrosarcoma involving the cavernous sinus and the brain stem. During the operation to remove the tumor, the intracavernous ICA was found to be severely invaded, and it ruptured. An emergent radial artery bypass graft was performed and the tumor was removed totally in two operations. She lost her vision (already impaired severely) but made an otherwise uneventful recovery. She remains tumor free after 3 years.
Case Study 4 Reconstruction of Vein of Labb´e
Transverse and sigmoid sinuses Figure 24 Operative picture of the completed anastomosis. The arrow points to the saphenous vein graft (SVG).
A 58-year-old woman presented with facial pain and numbness caused by a right petroclival meningioma extending into the Meckel’s cave and the cavernous sinus with mild brain stem compression (Fig. 20). The venous phase of the angiogram showed a prominent vein of Labb´e (Fig. 21). The tumor was removed by a transpetrosal retrolabyrinthine approach. However, early in the operation, an aberrant vein of Labb´e, draining the entire temporal lobe and draining into a dural sinus anterior to the transverse sinus, was damaged.
Chapter 15: Cerebrovascular Management in Skull Base Tumors
(A)
221
(B)
Figure 25 (A, B) Three-dimensional computed tomographic angiogram of Case 4 showing a patent graft and sinus.
(A)
(B)
Figure 26 Preoperative axial (A) and sagittal (B) enhanced MRI scan showing the lesion filling the right jugular foramen, petrous bone, and upper cervical area, with encasement and slight narrowing of the petrous ICA.
(A)
Figure 27 (A) Preoperative right carotid angiogram reveals poor connection between the two transverse sinuses and the dominant right sinus. (Part B located on page 220).
222
Natarajan et al.
(B)
Figure 27 (B) Preoperative left carotid angiogram shows partial filling of the right carotid circulation and a stenotic but patent right jugular bulb.
Because significant temporal lobe swelling was noted, the vein of Labb´e was reconstructed with a short saphenous vein graft from the vein to the sigmoid sinus (Fig. 22). A special technique of venous attachment to the sigmoid sinus was used without significantly interrupting flow by placing the attaching sutures first and then cutting the hole into the sinus before tying the sutures (Figs. 23 and 24). The tumor was seen to involve the trigeminal fascicles severely. The temporal lobe swelling resolved postoperatively, and the patient recovered well. Postoperatively, the patient had partial sixth cranial nerve palsy and diminished sensation in the V1 and V2 region and absent corneal reflex. Postoperative MRV and threedimensional computed tomographic angiogram showed patency of the graft (Fig. 25). After a follow-up of 2 years, there was no tumor recurrence. The sixth cranial nerve palsy disappeared totally, but trigeminal loss persisted. The venous reconstruction was felt to be important in this patient in avoiding major problems.
Case Study 5 Vein Graft from IJV to Sigmoid Sinus
Figure 28 Illustration showing the saphenous vein graft, CN VII, the tumor bed, and the petrous ICA.
Figure 29 Postoperative angiography shows excellent patency of the vein graft (between arrows).
A 62-year-old man presented with worsening seizure, right sided tinnitus, and swallowing disturbances. On examination, he had 20% hearing loss in his right ear, was unable to perform tandem walk, and had nystagmus. MRI scan showed a highly vascular glomus jugulare tumor in the jugular foramen filling the region of right jugular bulb (Fig. 26) without occluding the sigmoid sinus and internal jugular vein (Fig. 27). The tumor encased the petrosal segment of the ICA bowing it forward. The sigmoid and transverse sinuses were larger on the right side but were subtotally occluded by the tumor. There was a good communication between the two
Chapter 15: Cerebrovascular Management in Skull Base Tumors
sinuses at the torcular Herophili (Fig. 27). Therefore, it was felt to be safe to occlude the sinus during tumor resection. At operation, the intrasinus pressure was measured prior to sinus occlusion and excision of the tumor. Although the intrasinus pressure was initially unchanged after occlusion, the pressure increased steadily during the operation in the range of 35 to 40 torr. Because of this, reconstruction of the sinus was elected and performed with a saphenous vein graft (Fig. 28). A 5 cm long vein graft was sutured from the sigmoid sinus to the internal jugular vein. Patency of the graft was verified by intra-arterial and magnetic resonance angiography (Fig. 29). Postoperatively the patient had communicating hydrocephalus for which lumboperitoneal shunt was done. The patient had transient, postoperative facial nerve palsy, which recovered completely. After a follow up of 21 months, the patient was seen to have very mild impairment of tandem gait. There was no recurrence of the tumor and there was no other neurologic deficit.
REFERENCES 1. Mohit AA, Sekhar LN, Natarajan SK, et al. High-flow bypass grafts in the management of complex intracranial aneurysms. Neurosurgery. 2007;60:ONS105–122; discussion ONS122–103. 2. Natarajan SK, Sekhar LN, Schessel D, et al. Petroclival meningiomas: Multimodality treatment and outcomes at long-term follow-up. Neurosurgery. 2007;60:965–979; discussion 979–981. 3. Sekhar LN, Bucur SD, Bank WO, et al. Venous and arterial bypass grafts for difficult tumors, aneurysms, and occlusive vascular lesions: Evolution of surgical treatment and improved graft results. Neurosurgery. 1999;44:1207–1223; discussion 1223–1204. 4. Sekhar LN, Kalavakonda C. Cerebral revascularization for aneurysms and tumors. Neurosurgery. 2002;50:321–331. 5. Sekhar LN, Tzortzidis FN, Bejjani GK, et al. Saphenous vein graft bypass of the sigmoid sinus and jugular bulb during the removal of glomus jugulare tumors. Report of two cases. J Neurosurg. 1997;86:1036–1041. 6. Tzortzidis F, Elahi F, Wright D, et al. Patient outcome at long-term follow-up after aggressive microsurgical resection of cranial base chordomas. Neurosurgery. 2006;59:230–237; discussion 230–237.
223
7. Tzortzidis F, Elahi F, Wright DC, et al. Patient outcome at longterm follow-up after aggressive microsurgical resection of cranial base chondrosarcomas. Neurosurgery. 2006;58:1090–1098; discussion 1090–1098. 8. Larson JJ, Tew JM Jr, Tomsick TA, et al. Treatment of aneurysms of the internal carotid artery by intravascular balloon occlusion: Long-term follow-up of 58 patients. Neurosurgery. 1995;36:26– 30; discussion 30. 9. McIvor NP, Willinsky RA, TerBrugge KG, et al. Validity of test occlusion studies prior to internal carotid artery sacrifice. Head Neck. 1994;16:11–16. 10. Origitano TC, Al-Mefty O, Leonetti JP, et al. Vascular considerations and complications in cranial base surgery. Neurosurgery. 1994;35:351–362; discussion 362–353. 11. Sekhar LN, Patel SJ. Permanent occlusion of the internal carotid artery during skull-base and vascular surgery: Is it really safe? Am J Otol. 1993;14:421–422. 12. Evans JJ, Sekhar LN, Rak R, et al. Bypass grafting and revascularization in the management of posterior circulation aneurysms. Neurosurgery. 2004;55:1036–1049. 13. Guppy KH, Origitano TC, Reichman OH, et al. Venous drainage of the inferolateral temporal lobe in relationship to transtemporal/transtentorial approaches to the cranial base. Neurosurgery. 1997;41:615–619; discussion 619–620. 14. Koperna T, Tschabitscher M, Knosp E. The termination of the vein of “Labbe” and its microsurgical significance. Acta Neurochir (Wien). 1992;118:172–175. 15. Kyoshima K, Oikawa S, Kobayashi S. Preservation of large bridging veins of the cranial base: Technical note. Neurosurgery. 2001;48:447–449. 16. Lustig LR, Jackler RK. The vulnerability of the vein of labbe during combined craniotomies of the middle and posterior fossae. Skull Base Surg. 1998;8:1–9. 17. Morita A, Sekhar LN. Reconstruction of the vein of Labbe by using a short saphenous vein bypass graft. Technical note. J Neurosurg. 1998;89:671–675. 18. Sakata K, Al-Mefty O, Yamamoto I. Venous consideration in petrosal approach: Microsurgical anatomy of the temporal bridging vein. Neurosurgery. 2000;47:153–160; discussion 160– 151. 19. Bejjani GK, Sekhar LN, Yost AM, et al. Vasospasm after cranial base tumor resection: Pathogenesis, diagnosis, and therapy. Surg Neurol. 1999;52:577–583; discussion 583–574.
Section 2 Site-Specific Considerations
16 Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base Ehab Y. Hanna, Michael Kupferman, and Franco DeMonte
The roof of the nasal cavity is narrow from side to side, and slopes downward (at about a 30-degree angle) from front to back. The cribriform plate, which transmits the filaments of the olfactory nerve, forms the roof of the nasal cavity medial to the superior attachment of the middle turbinate. Lateral to the middle turbinate the fovea ethmoidalis forms the roof of the ethmoid sinuses. Careful assessment of the anatomy of the nasal roof, especially the relationship of the cribriform plate to the fovea ethmoidalis is critical in avoiding a cerebrospinal fluid (CSF) leak during transnasal surgery in this region. The cribriform plate is usually at a slightly lower horizontal plane than the fovea ethmoidalis forming a shallow olfactory groove. This configuration is described as Keros type I (Fig. 4). However, the cribriform plate may be moderately or significantly lower than the fovea ethmoidalis, resulting in a medium (Keros type II) or deep (Keros type III) olfactory groove. The topography of the roof may also be asymmetrical (1). The floor of the nasal cavity is concave from side to side and almost horizontal anteroposteriorly. The palatine process of the maxilla forms the anterior three-fourths, and the horizontal process of the palatine bone forms the posterior fourth of the nasal floor [Fig. 2(B)]. The majority of the nasal cavity is lined by pseudostratified ciliated columnar epithelium, which contains mucous and serous glands (respiratory epithelium). Specialized olfactory epithelium lines the most superior portion of the nasal cavity, and has direct connections with the olfactory tracts through openings in the cribriform plate. The arteries of the nasal cavities are the anterior and posterior ethmoidal branches of the ophthalmic artery, which supply the ethmoid and frontal sinuses and roof of the nose. The sphenopalatine artery supplies the mucous membrane covering the lateral nasal wall. The septal branch of the superior labial artery supplies the anteroinferior septum. The veins form a close cavernous plexus beneath the mucous membrane. This plexus is especially well marked over the lower part of the septum and over the middle and inferior turbinates. Venous drainage follows a pattern similar to arterial supply. The lymphatic drainage from the anterior part of the nasal cavity, similar to that of the external nose, is to the submandibular group of lymph nodes (Level I). Lymphatics from the posterior two-thirds of the nasal cavities and from the paranasal sinuses drain to the upper jugular (Level II) and retropharyngeal lymph nodes. The sensory nerves of the nasal cavity transmit either somato-autonomic or olfactory sensation. Somato-autonomic nerves include the nasociliary branch of the ophthalmic, which supplies the anterior septum and lateral wall. The anterior alveolar nerve, a branch of the maxillary (V2),
SURGICAL ANATOMY Nasal Cavity The nasal cavity is bounded by the bony pyriform aperture and the external framework of the nose (Fig. 1). The nasal cavity opens anteriorly through the skin-lined nasal vestibule into the nares, and communicates posteriorly through the choanae with the nasopharynx [Fig. 2(A)]. The nasal cavity is divided in the midline by the nasal septum, which includes both cartilaginous and bony components [Fig. 2(B)]. The cartilage of the septum is quadrilateral in shape, and is thicker at its margins than at its center. Its anterior margin is connected with the nasal bones, and is continuous with the anterior margins of the lateral cartilages; below, it is connected to the medial crura of the greater alar cartilages by fibrous tissue (Fig. 1). Its posterior margin is connected with the perpendicular plate of the ethmoid; its inferior margin is connected with the vomer and the palatine process of the maxilla. On the lateral nasal wall are the superior, middle, and inferior nasal turbinates, and below and lateral to each turbinate (concha) is the corresponding nasal passage or meatus [Fig. 3(A) and 3(B)]. Above the superior turbinate is a narrow recess, the sphenoethmoidal recess, into which the sphenoid sinus opens. The superior meatus is a short oblique passage extending about halfway along the upper border of the middle turbinate; the posterior ethmoid cells open into the front part of this meatus. The middle meatus is below and lateral to the middle turbinate. The anatomy of the middle meatus is fully displayed by removing the middle turbinate [Fig. 3(A) and 3(B)]. The bulla ethmoidalis is the most prominent anterior ethmoid air cell. The hiatus semilunaris is a curved cleft lying below and in front of the bulla ethmoidalis. It is bounded inferiorly by the sharp concave margin of the uncinate process of the ethmoid bone, and leads into a curved channel, the infundibulum, bounded above by the bulla ethmoidalis and below by the lateral surface of the uncinate process of the ethmoid. The anterior ethmoid air cells open into the front part of the infundibulum. The frontal sinus drains through the nasofrontal duct, which, in approximately 50% of subjects, will also drain into the infundibulum, but when the anterior end of the uncinate process fuses with the front part of the bulla, this continuity is interrupted and the nasofrontal duct then opens directly into the anterior end of the middle meatus. Below the bulla ethmoidalis, and partly hidden by the inferior end of the uncinate process, is the ostium of the maxillary sinus. An accessory ostium from the maxillary sinus is frequently present below the posterior end of the middle nasal concha. The inferior meatus is below and lateral to the inferior nasal turbinate. The nasolacrimal duct opens into the inferior meatus under cover of the anterior part of the inferior turbinate. 227
228
Hanna et al.
Anterolateral view Frontal bone Nasal bones Frontal process of maxilla
Inferior view Major alar cartilage Lateral crus
Medial crus
Lateral process of septal nasal cartilages Septal cartilage Minor alar cartilage Accessory nasal cartilage
Major alar cartilage
Lateral crus Medial crus
Septal nasal cartilage Anterior nasal spine of maxilla Alar fibrofatty tissue
Alar fibrofatty tissue Septal nasal cartilage
Infraorbital foramen
provides sensory innervation to the inferior meatus and inferior turbinate. The nasopalatine nerve supplies the middle of the septum. The anterior palatine nerve supplies the lower nasal branches to the middle and inferior turbinates. The nerve of the pterygoid canal (vidian) and the nasal branches from the sphenopalatine ganglion supply the upper and posterior septum and superior turbinate. The olfactory nerve fibers arise from the bipolar olfactory cells and unite in fasciculi, which form a plexus beneath the mucous membrane and then ascend passing into the skull through the foramina in the cribriform plate. Intracranially, olfactory nerve fibers enter the undersurface of the olfactory bulb, in which they ramify and form synapses with the dendrites of the mitral cells of the olfactory tract.
Anterior nasal spine of maxilla Intermaxillary suture
Figure 1 Anatomy of the external nose. Source: From: Netter’s Atlas of Human Anatomy, Elsevier Inc.
the zygomatic process. Its roof or orbital wall is frequently ridged by the infraorbital canal, while its floor is formed by the alveolar process of the maxilla and is usually 1 to 10 mm below the level of the floor of the nose (Fig. 6). Projecting into the floor are several conical elevations corresponding with the roots of the first and second molar teeth, and in some cases the floor is perforated by one or more of these roots. The natural ostium of the maxillary sinus is partially covered by the uncinate process and communicates with the lower part of the hiatus semilunaris of the lateral nasal wall (Figs. 3 and 5). An accessory ostium is frequently seen in, or immediately behind, the hiatus. The maxillary sinus appears as a shallow groove on the medial surface of the bone about the fourth month of fetal life, but does not reach its full size until after the second dentition.
Maxillary Sinus The maxillary sinus (antrum of Highmore), the largest of the accessory sinuses of the nose, is a pyramidal cavity in the body of the maxilla (Figs. 5 and 6). Its base is formed by the lateral wall of the nasal cavity, and its apex extends into
Ethmoid Sinus The ethmoidal air cells consist of numerous thin-walled cavities situated in the ethmoidal labyrinth and bounded by the frontal, maxillary, lacrimal, sphenoid, and palatine bones.
Medial Wall of Nasal Cavity (Nasal Septum)
Frontal sinus Sphenoidal sinus Nasal septum
Falx cerebri Cribriform plate of ethmoid bon. Dura mater Sella turrica
Frontal bone
Choanae Lateral process of septal nasal cartilage
Nasal bone Vomer
Major alar cartilage Nasal vestibute Anterior nasal spine Incisive canal
Squamous part Sinus Nasal spine
Vomerine grove (for nasopalatine nerve and vessels Septal cartilage Major alar cartilage (medial crus)
Oral cavity Tongue Soft palate
Anterior nasal spine Maxilla
Nasal crest Incisive canal Palatine process
Crista galli Cribriform plate Perpendicular plate
Ethmoid bone Crest Body Sphenoidal sinus Medial, Lateral Plates of pterygoid process
Spenoid bone
Basilar part of occipital bone Pharyngeal tubercle Perpendicular plate Nasal crest Posterior nasal spine Horizontal plate Lesser palatine foramen
Palatine bone
Greater palatine foramen
Figure 2 (A) Anatomy of the nasal septum. (B) Skeletal framework of the nasal septum. Source: From Netter’s Atlas of Human Anatomy, Elsevier Inc.
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
puberty. The frontal sinus is lined with respiratory epithelium and drains into the anterior part of the corresponding middle meatus of the nose through the nasofrontal duct, which traverses the anterior part of the labyrinth of the ethmoid. The soft tissues of the forehead are located anteriorly, the orbits are located inferiorly, and the anterior cranial fossa is located posteriorly (Fig. 3). Blood and neural supply is from the supraorbital and supratrochlear neurovascular bundles.
They lie in the upper part of the nasal cavity between the orbits (Fig. 6). The ethmoid sinuses are separated from the orbital cavity by a thin bony plate, the lamina papyracea. On either side they are arranged in three groups: anterior, middle, and posterior. The anterior and middle groups open into the middle meatus of the nose, the former by way of the infundibulum and the latter on or above the bulla ethmoidalis (Fig. 3). The posterior cells open into the superior meatus under cover of the superior nasal concha. Sometimes one or more ethmoid air cells extend over the orbital cavity (supraorbital ethmoid cells) or the optic nerve (Onodi cell). The ethmoidal cells begin to develop during fetal life.
Sphenoid Sinus The sphenoid sinus begins at the most posterior and superior portion of the nasal cavity (Fig. 3). This midline structure, which is contained within the body of the sphenoid bone, is irregular and often has an eccentrically located intersinus septum. When exceptionally large, the sphenoid sinus may extend into the roots of the pterygoid processes or great wings, and may pneumatize the basilar part of the occipital bone. The sphenoid sinus ostium is located on the anterior wall of the sinus and communicates directly with the sphenoethmoidal recess above and medial to the superior
Frontal Sinus The paired frontal sinuses appear to be outgrowths from the most anterior ethmoidal air cells. They are situated behind the superciliary arches, are rarely symmetrical, and the septum between them frequently deviates to one or the other side of the midline. Absent at birth, the frontal sinuses are generally fairly well developed between the seventh and eighth years, but only reach their full size after
Lateral Wall of Nasal Cavity Sphenoethmoidal recess Opening of sphenoidal sinus Hypophysis (pituitary gland) in sella turcica Sphenoidal sinus Phyaryngeal tonsil (adenoid if enlarged) Basilar part of occipital bone Pharyngeal raphe Choana Torus tubarius Opening of pharyngotympanic (auditory) tube Pharyngeal recess Horizontal plate of palatine bone Soft palate
Frontal sinus Superior nasal concha Superior nasal meatus Middle nasal concha Agger nasi Atrium of middle nasal meatus Middle nasal meatus Inferior nasal concha (turbinate) Limen nasi Nasal vestibute Inferior nasal mealus Palatine process of maxilla Incisive canal Tongue
Middle nasal concha Middle nasal meatus Bulging septum Airway to nasopoharynx Inferior nasal concha Inferior nasal meatus Floor of nasal cavity Speculum view Frontal sinus Probe passing from semilunar hiatus into frontal sinus via frontonasal duct
Cribriform plate of ethmoid bone Probe in opening of sphenoidal sinus Sphenoidal sinus Superior nasal meatus with opening of posterior ethmoidal cells Basilar part of occipital bone Torus tubarius Opening of phyaryngotympanic (auditory) tube Anterior arch of atlas (C.I vertebra)
Middle nasal concha (cut surface) Ethmoidal bulla Openings of middle ethmoidal cells Semilunar hiatus with opening of anterior ethmoidal cells Uncinate process Inferior nasal concha (cut surface) Opening of nasolacrimal duct Inferior nasal meatus
Figure 3
229
Opening of maxillary sinus
Dens of axis (C2 vertebra)
(A) Anatomy of the lateral nasal wall. Removal of the middle turbinate demonstrates the anatomy of the middle meatus (lower image).
230
Hanna et al.
Lacrimal bone Middle nasal concha Uncinate process Ethmoid bone Cribriform plate Superior nasal concha Highest nasal concha
Squamous part Sinus Nasal spine
Frontal bone
Sphenoethmoidal recess
Nasal bone
Sphenopalatine foramen
Agger nasi
Basilar part of occipital bone
Lateral process of nasal septal cartilage Major alar cartilage Alar fibrofatty tissue
Frontal process Anterior nasal spine Maxilla
Incisive canal Palatine process Alveolar process
Inferior nasal concha
Sphenoidal sinus Medical and Lateral Sphenoid bone plates of pterygoid process Pterygoid hamulus Sphenoidal process Orbital process Posterior nasal spine Palatine bone Perpendicular plate Horizontal plate
Ethmoidal bulla Frontal sinus Opening of frontonasal canal
Opening of middle ethmoidal cells Superior nasal concha (cut away) Openings of posterior ethmoidal cells
Middle nasal concha (cut away) Openings of sphenoidal sinus Infundibulum leading to frontonasal canal Semilunar hiatus (osteomeatal unit) with opening of anterior ethmoidal air cells
Openings into maxillary sinus Sphenopalatine foramen
Uncinate process Inferior nasal concha (cut away) Opening of nasolacrimal canal
Ethmoidal process of inferior nasal concha Lesser palatine foramen Greater palatine foramen
Figure 3 (B) Skeletal framework of the lateral nasal wall. Removal of the middle turbinate demonstrates the anatomy of the middle meatus (lower image). Source: From Netter’s Atlas of Human Anatomy, Elsevier Inc.
turbinate (Fig. 3). The sphenoid sinuses are present as minute cavities at birth, but their main development takes place after puberty. The posterosuperior wall of the sphenoid sinus displays the forward convexity caused by the floor of the sella turcica, which contains the pituitary gland. The optic nerve and the internal carotid artery are closely related to the superolateral wall of the sphenoid sinus, and their bony canals may be dehiscent within the sinus cavity (Fig. 7). Vascular and neural supplies come from the sphenopalatine and posterior ethmoidal arteries and the branches of the sphenopalatine ganglion, respectively.
Infratemporal Fossa The infratemporal fossa is an irregularly shaped space, situated below and medial to the zygomatic arch. It is bounded, anteriorly by the posterior surface of the maxilla, superiorly by the greater wing of the sphenoid and by the undersurface of the squamous portion of the temporal bone medially by the lateral pterygoid plate; and laterally by the
ramus of the mandible. It contains the inferior aspect of the temporalis muscle, and the medial and lateral pterygoid muscles (Fig. 8). It also contains branches of the internal maxillary vessels including the middle meningeal artery, and the mandibular (V3) nerves including the lingual, inferior alveolar, and auriculotemporal nerves. The foramen ovale and foramen spinosum open on its roof, and the alveolar canals on its anterior wall. The inferior orbital and pterygomaxillary fissures communicate with and may act as routes of spread of cancer to the infratemporal fossa. The infratemporal fossa also contains the upper carotid sheath, including the internal carotid artery, internal jugular vein and the last four cranial nerves (Fig. 8).
Pterygopalatine Fossa The pterygopalatine fossa is a small, triangular space situated behind the maxillary sinus, in front of the pterygoid plates, and beneath the apex of the orbit. This fossa communicates with the orbit by the inferior orbital fissure, with the nasal
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
231
foramina: the anterior, situated about the middle of the lateral margin of the olfactory groove, transmits the anterior ethmoidal vessels and the nasociliary nerve; the nerve runs in a groove along the lateral edge of the cribriform plate; and the posterior ethmoidal foramen opens at the back part of this margin under cover of the projecting lamina of the sphenoid, and transmits the posterior ethmoidal vessels and nerve. More laterally, the cranial floor forms the orbital roof and supports the frontal lobes of the cerebrum. Further back in the middle is the planum sphenoidale forming the roof of the sphenoid sinus, and the anterior margin of the chiasmatic groove, running laterally on either side to the upper margin of the optic foramen (Fig. 9).
Orbit
Figure 4 Anatomy of the ethmoid roof and lateral lamella of the cribriform plate: (A) Keros type I, (B) Keros type II, (C) Keros type III, and (D) asymmetrical ethmoid roof. Note that the right lateral lamella of the cribriform plate is very thin and long and is obliquely oriented including much of the right ethmoid roof. Abbreviations: OP, Orbital plate of frontal bone; LCPL, Lateral cribriform plate lamella; EB, Ethmoid bulla.
cavity by the sphenopalatine foramen, and with the infratemporal fossa by the pterygomaxillary fissure (Fig. 5). Five foramina open into it. Of these, three are on the posterior walls, which are the foramen rotundum, the pterygoid canal, and the pharyngeal canal, in this order downward and medial. On the medial wall is the sphenopalatine foramen, and below is the superior orifice of the pterygopalatine canal (Fig. 5). The fossa contains the maxillary nerve, the sphenopalatine ganglion, and the terminal part of the internal maxillary artery. The fissures and foramina of the pterygopalatine fossa serve as “highways” for spread of cancer from the sinonasal region to the orbit, infratemporal fossa, and cranial base.
Anterior Cranial Fossa The floor of the anterior fossa is formed by the orbital plates of the frontal bone, the cribriform plate of the ethmoid, and the lesser wings and front part of the body of the sphenoid. In the midline, it presents, from anterior to posterior, the frontal crest for the attachment of the falx cerebri; the foramen cecum, which usually transmits a small vein from the nasal cavity to the superior sagittal sinus; the crista galli, the free margin of which affords attachment to the falx cerebri (Fig. 9). On either side of the crista galli is the olfactory groove formed by the cribriform plate, which supports the olfactory bulb and presents foramina for the transmission of the olfactory nerves. Lateral to either olfactory groove are the internal openings of the anterior and posterior ethmoidal
The orbits are two quadrilateral pyramidal cavities, their bases being directed forward and lateral, and their apices backward and medial, so that their long axes diverge at a 45degree angle and, if continued backward, would meet over the body of the sphenoid. The orbit is anatomically defined by seven bones (Fig. 10): frontal, zygomatic, maxillary, lacrimal, ethmoid, sphenoid and palatine; and by the orbital septum which originates at the arcus marginalis, fusing with the levator aponeurosis above and the capsulopalpebral fascia below. It is bounded by the ethmoid and sphenoid sinuses at its medial aspect, the frontal sinus superomedially, the cranial vault superiorly and posteriorly, the temporal fossa laterally, and the maxillary sinus inferiorly. Each orbital cavity has a roof, a floor, a medial and a lateral wall, a base, and an apex. The roof is formed anteriorly by the orbital plate of the frontal bone, and posteriorly by the lesser wing of the sphenoid. It presents medially the trochlear fovea for the attachment of the cartilaginous pulley of the superior oblique muscle and laterally the lacrimal fossa for the lacrimal gland. The floor is formed mainly by the orbital surface of the maxilla; anteriorly and laterally, by the orbital process of the zygomatic bone; and posteriorly and medially, to a small extent, by the orbital process of the palatine bone. At its medial angle is the superior opening of the nasolacrimal canal, immediately to the lateral side of which is a depression for the origin of the inferior oblique muscle. Running anteriorly near the middle of the floor is the infraorbital canal, ending anterior to the maxilla in the infraorbital foramen and transmitting the infraorbital nerve and vessels. The medial wall is formed from anteriorly to posteriorly by the frontal process of the maxilla, the lacrimal bone, the lamina papyracea of the ethmoid, and a small part of the body of the sphenoid anterior to the optic foramen. Anteroinferiorly, the lacrimal sac is situated between the anterior and posterior lacrimal crests, at the junction between the medial wall and the floor. The lacrimal part of the orbicularis oculi arises from the posterior lacrimal crest. At the junction of the medial wall and the roof, the frontoethmoidal suture presents the anterior and posterior ethmoidal foramina, the former transmitting the nasociliary nerve and anterior ethmoidal vessels and the latter the posterior ethmoidal nerve and vessels. These foramina indicate the level of the cranial base within the orbit. The lateral wall is formed by the orbital process of the zygomatic and the orbital surface of the greater wing of the sphenoid. On the orbital process of the zygomatic bone are the orbital tubercle (Whitnall) and the orifices of one or two canals, which transmit the branches of the zygomatic nerve. Between the roof and the lateral wall, near the apex of the orbit, is the superior orbital fissure (SOF). Through this fissure the oculomotor, the trochlear, the ophthalmic division
232
Hanna et al.
Frontal sinus
Posterior ethmoidal foramen Orbital process of palatine Optic foramen
Anterior ethmoidal foramen
Sphenopalatine foramen Sella turcica Probe in foramen rotundum
Fossa for lacrimal sac Uncinate process of ethmoid Openings of maxillary sinus Inferior nasal concha
Probe in pterygoid canal Probe in pterygopalatine canal Palatine bone Lateral pterygoid plate
Pyramidal process of palatine
Figure 5
Anatomy of the maxillary sinus (lateral wall removed). Source: From Ref. 22.
of the trigeminal (V1), and the abducent nerves enter the orbital cavity, also some filaments from the cavernous plexus of the sympathetic and the orbital branches of the middle meningeal artery. Passing posteriorly through the fissure are the ophthalmic vein and the recurrent branch from the lacrimal artery to the dura mater. The lateral wall and the floor are separated posteriorly by the inferior orbital fissure, which transmits the maxillary nerve (V2) and its zygomatic branch, the infraorbital vessels, and the ascending branches from the sphenopalatine ganglion. The base of the orbit (orbital rim), quadrilateral in shape, is formed superiorly by the supraorbital arch of the frontal bone, in which is the supraorbital notch or foramen for the passage of the supraorbital vessels and nerve; inferiorly by the zygomatic bone and maxilla, united by the zygomatico-maxillary suture; laterally by the zygomatic bone and the zygomatic process of the frontal joined by the zygomatico-frontal suture; medially by the frontal bone and the frontal process of the maxilla united by the frontomaxillary suture. The apex is situated in the posterior aspect of the orbit. The optic foramen is a short, cylindrical canal, through which passes the optic nerve and ophthalmic artery. The extraocular muscles—four rectus muscles and two obliques—effect movement of the eye. The third cranial nerve innervates all but the lateral rectus and the superior oblique
muscles, which are innervated by the sixth and fourth cranial nerves, respectively. The rectus muscles originate at the annulus of Zinn and insert on the globe forming a muscle cone, which is the central anatomic space in the orbit. The lacrimal system is composed of secretory and drainage systems. Secretory glands—the glands of Moll, Kraus, and Wolfring—may be found along the margin of the eyelid. The lacrimal gland with its palpebral and orbital lobes is located in the superotemporal orbit (Fig. 11). The lacrimal drainage system, located in the inferonasal orbit, is represented by the puncta, canaliculi, lacrimal sac, and nasolacrimal duct. Tumor involvement of the lacrimal system may present with epiphora. The skin of the eyelid is continuous with the palpebral and bulbar conjunctivae, which are, in turn, contiguous with the globe. Each of these epithelial surfaces represents a potential site of origin for cancer.
REGIONAL PATHOLOGY AND DIFFERENTIAL DIAGNOSIS Tumors of the Nasal Cavity and Paranasal Sinuses The mucosal lining of the nose—the Schneiderian membrane— is derived from ectoderm. This is uniquely different from the mucosa of the rest of the upper respiratory tract, which is derived from endoderm. Olfactory neuroepithelium lines
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
233
Coronal section Olfactory bulbs
Celebral falx Brain
Frontal sinus
Nasal cavities
Orbital fat
Nasal septum Ethmoidal cells Middle nasal concha
Opening of maxillary sinus
Middle nasal meatus
Infraorbital Zygomatic
Maxillary sinus
Alveolar
Inferior nasal meatus
Recesses of maxillary sinus
Buccinator muscle Alveolar process of maxilla
Inferior nasal concha
Body of tongue Hard palate Sublingual gland Oral cavity Mandible (body)
Sagittal section Ethmoidal cells Frontal sinus
Opening of sphenoidal sinus
Opening of frontonasal duct Sphenoidal sinus
Semilunar hiatus
Middle nasal concha (cut away)
Uncinate process Opening of maxillary sinus Inferior nasal concha
Figure 6 Anatomy of the ethmoid and maxillary sinuses. (Coronal section). Source: From Netter’s Atlas of Human Anatomy, Elsevier Inc.
the superior portion of the nasal cavity and the roof of the nose, and gives rise to neuroectodermal tumors such as olfactory neuroblastoma and neuroendocrine carcinoma (2–4). The sinonasal epithelium also has minor salivary glands, and gives rise to salivary gland tumors such as adenoidcystic carcinoma and mucoepidermoid carcinoma (5–8). However, the most common epithelial neoplasms of the sinonasal tract are those arising from “metaplastic” squamous epithelium, namely, squamous cell carcinoma, and those originating from the seromucinous glands of the mucosal lining, collectively known as adenocarcinomas (9). The unique histology of this region is reflected in the histogenesis of a complex variety of epithelial and nonepithelial tumors (Table 1). These tumors have a wide range of biologic behavior, and a few arise exclusively in the sinonasal tract (e.g., inverted papilloma, olfactory neuroblastoma). Nonepithelial tumors are similar to those in other regions in the head and neck. The details of some of these tumors, such as squamous and nonsquamous carcinoma, melanoma, esthesioneuroblastoma, sarcomas, and angiofibromas and fibro-osseous lesion, are discussed in more detail in other chapters in section three of the book. Overall, sinonasal cancer accounts for about 1% of all malignancies and approximately 3% of cancer of the head
and neck. There is a male predominance [Fig. 12(A)], with a strong predilection for Caucasians [Fig. 12(B)]. The majority of patients are over 50 years of age at the time of diagnosis [Fig. 12(C)]. The most common malignant tumor of the nasal cavity and paranasal sinuses is squamous cell carcinoma [Fig. 12(D)]. Although the maxillary antrum is the most commonly involved sinus [Fig. 12(E)], anterior skull base invasion is most frequently encountered with malignant neoplasms of the nasal cavity and ethmoid sinus. Upward extension of these neoplasms toward the cribriform plate or fovea ethmoidalis is not uncommon, and heralds intracranial extension (10). Primary carcinoma of the frontal sinus is uncommon, and those arising in the sphenoid sinus are rare (11). Unfortunately, despite significant improvement in diagnostic techniques such as nasal endoscopy and high-resolution imaging, most patients present with advanced stage disease [Fig. 12(F)].
Tumors of the Orbit The majority of malignant tumors involving the orbit represent direct extension of tumors of the sinonasal tract. Cancers arising primarily within the orbit are less common, and may be classified broadly into pediatric and adult groups. Further
234
Hanna et al.
(A)
(B)
Figure 7 (A) Cadaver dissection of the sphenoid sinus (SS). The sinus is located in the midline superior to the nasopharynx (NP). The sella turcica (ST) forms a convexity in the posterosuperior wall. The internal carotid artery (arrow) courses through the lateral wall of the sinus and is related superiorly to the optic nerve (ON). (B) Endoscopic view of the left sphenoid sinus. Note the internal carotid artery (ICA) and optic nerve (ON) impressions on the lateral and superior walls. A bony septum within the sinus inserts into the optico-carotid recess (OCR). Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
(A)
(B)
(C)
Figure 8 Anatomy of the infratemporal fossa. (A) Inferior view. (B) Lateral view. (C) Anterior view. Abbreviations: TM, temporalis muscle; MM, masseter muscle; MPM, medial pterygoid muscle; LPM, lateral pterygoid muscle; IMA, internal maxillary artery; MMA, middle meningeal artery; ICA, internal carotid artery; IJV, internal jugular vein; LN, lingual nerve; IAN, inferior alveolar nerve; ATN, auriculotemporal nerve; V3, third division of the trigeminal nerve; ET, Eustachian tube; CN, cranial nerve (VII, IX, X, XI, and XII).
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
235
Figure 9 The floor of the anterior cranial fossa. The cribriform plate (CP) is characterized by the presence of foramina for the olfactory nerves on each side of the crista galli (CG), which is seen in the midline. Lateral to the cribriform plate (CP) is the ethmoidal roof, and even more lateral the roof of the orbit. Posterior to the cribriform plate is the planum sphenoidale (PS). The optic nerves (ON) form the optic chiasm behind the planum sphenoidale. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
subclassification may be according to site of origin, histologic type, or both. The most common intraocular tumor seen in children is Retinoblastoma, which usually presents by the third year of age. Other common tumors in the orbit in children include rhabdomyosarcoma, neuroblastoma, lymphoma, and leukemia. Rhabdomyosarcoma is the most common primary cancer and Neuroblastoma is the most common metastatic cancer to the orbit in children. Granulocytic sarcoma as a primary orbital neoplasm may precede or follow systemic leukemia. Primarily found in children with myelogenous leukemia, this tumor rarely occurs in adults (1). In adults, approximately 65% of orbital tumors are malignant. The most common benign tumors are vascular malformations and pleomorphic adenoma of the lacrimal gland. Malignant tumors of the lacrimal gland are most commonly lymphomas or tumors of salivary gland origin. Overall, lymphoma is the most common tumor of the orbit in adults. The main differential diagnosis is lymphoid hyperplasia and orbital pseudotumor. Malignant salivary gland tumors of the lacrimal gland include adenoid cystic carcinoma, ma-
lignant mixed cell tumor, and mucoepidermoid carcinoma. Neoplasms of the lacrimal sac include squamous cell carcinoma, adenocarcinoma, transitional cell carcinoma, salivary gland carcinoma, and poorly differentiated carcinoma. Cancer of the skin of the lid includes basal cell carcinoma, squamous cell carcinoma, sebaceous cell carcinoma and malignant melanoma, any of which may invade the orbit. Tumors arising from the conjunctiva may also invade the orbit, including malignant melanoma, squamous carcinoma, and lymphoma. Choroidal melanoma is the most common of all intraocular malignancies, and is biologically distinct from conjunctival or cutaneous melanoma (1). Primary intraorbital malignancies in adults are rare. Malignant neurogenic tumors of the orbit are uncommon, but those of peripheral nerve sheath origin predominate. They most commonly represent malignant degeneration in patients with multiple benign neurofibromatosis. Other sarcomas that infrequently arise within the orbit include osteosarcoma, chondrosarcoma, malignant fibrous histiocytoma, hemangiopericytoma, and liposarcoma. Multiple myeloma may present in the orbit as a solitary plasmacytoma
236
Hanna et al.
Orbital surface of frontal bone Optic canal (foramen)
Posterior ethmoidal foramen Anterior ethmoidal foramen
Orbital surface of lesser wing of sphenoid bone Superior orbital fissure
Orbital plate of ethmoid bone
Orbital surface of greater wing of sphenoid bone Lacrimal bone Orbital surface of zygomatic bone
Fossa for lacrimal sac Inferior orbital fissure
Orbital surface of maxilla Infraorbital groove Orbital process of palatine bone
Figure 10 Anatomy of the right orbit. Source: From Ref. 22.
or the orbit may be involved as part of disseminated disease. Hematogenous metastasis to the orbit most commonly originates from a primary in the lung or prostate in males. In females, carcinoma of the breast is the most common source of metastasis to the orbit (1).
Frontal bone (cut away) Orbital part of lacrimal gland Palpebral part of lacrimal gland Excretory ducts of lacrimal gland Plica semilunaris and lacrimal lake Lacrimal caruncle Inferior lacrimal papilla and punctum Opening of nasolacrimal duct
Superior lacrimal papilla and punctum Lacrimal canaliculi Lacrimal sac Nasolacrimal duct Middle nasal concha Nasal cavity Inferior nasal concha (cut) Inferior nasal meatus
Figure 11 Anatomy of the lacrimal system. Source: From: Netter’s Atlas of Human Anatomy, Elsevier Inc.
CLINICAL ASSESSMENT Clinical evaluation of patients with cancer of the nasal cavity, paranasal sinuses, and orbit should help to achieve three objectives: (1) establishment of the diagnosis, (2) determination of the extent of and stage disease, and (3) development of a plan for treatment. These objectives are usually achieved through a detailed history, comprehensive clinical examination of the head and neck, imaging, and biopsy.
History and Clinical Examination The signs and symptoms of early sinonasal tumors are very subtle and nonspecific. Early lesions are often completely asymptomatic or mimic more common benign conditions such as chronic sinusitis, allergy, or nasal polyposis. Since early detection of sinonasal tumors is probably the most important factor in improving prognosis, a high degree of suspicion is necessary to diagnose smaller lesions. Common symptoms include nasal obstruction, “sinus pressure” or pain, nasal discharge that may be bloody, anosmia, or epistaxis. Failure of these symptoms to respond to adequate medical therapy or the presence of unilateral signs and symptoms should alert the physician to the possibility of malignancy, and warrants further investigation by high-resolution
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base Table 1 Tumors of the Sinonasal Tract Benign Epithelial Papilloma Adenoma Dermoid Nonepithelial Fibroma Chondroma Osteoma Neurofibroma Hemangioma Lymphangioma Nasal glioma Intermediate Schneiderian papilloma Inverted Papillary Cylindrical Angiofibroma Ameloblastoma Fibrous dysplasia Ossifying fibroma Giant cell tumor Malignant Epithelial Squamous cell carcinoma Differentiated (well, moderately, poorly) Basaloid squamous Adenosquamous Nonsquamous cell carcinoma Adenoid cystic carcinoma Mucoepidermoid carcinoma Adenocarcinoma Neuroendocrine carcinoma Hyalinizing clear cell carcinoma Melanoma Olfactory neuroblastoma Sinonasal undifferentiated carcinoma Nonepithelial Chordoma Chondrosarcoma Osteogenic sarcoma Soft tissue sarcoma Fibrosarcoma Malignant fibrous histiocytoma Hemangiopericytoma Angiosarcoma Kaposi sarcoma Rhabdomyosarcoma Lymphoproliferative Lymphoma Polymorphic reticulosis Plasmacytoma Metastatic Source: From Ref. 1
imaging. Comprehensive examination of the nasal cavity should be done after topical decongestion and anesthesia using rigid or flexible endoscopy (Fig. 13). The presence of intranasal masses, ulcers, or areas of contact bleeding may indicate a malignant tumor. Although unilateral “polyps” may be inflammatory, they are more commonly neoplastic. Tumors may also present as submucosal masses without
237
changes in the mucosa, other than displacement. Any suspicious lesions should be biopsied; preferably after highresolution imaging has been obtained to avoid severe bleeding and/or CSF leak as discussed below. Extension of sinonasal tumors to adjacent structures renders the diagnosis obvious, but is a late manifestation of the disease. Soft tissue swelling of the face may indicate tumor extension through the anterior bony confines of the nose and sinuses (Fig. 14). Inferior extension toward the oral cavity may present with an ulcer or a submucosal mass in the palate or the alveolar ridge (Fig. 15). Middle ear effusion may indicate tumor involvement of the nasopharynx, Eustachian tube, pterygoid plates, or tensor veli palatini muscle. Extension to the skull base may lead to involvement of the cranial nerves leading anosmia, blurred vision, diplopia, or in hyposthesia along the branches of the trigeminal nerves. The presence of associated neck masses usually represents metastatic disease in the cervical lymph nodes. Orbital involvement is common in patients with cancer arising from the ethmoid, maxillary, frontal, and sphenoid sinuses in descending order of frequency. Less commonly, the orbit is involved by a primary tumor of the eye or its adnexa. Signs and symptoms of tumors in the orbit are usually due to mass effect or neuromuscular dysfunction. The patient may complain of proptosis, irregular shape of the eyelid, or blepharoptosis. Epiphora usually indicates involvement of the nasolacrimal duct (Fig. 11). Double vision may result from compression or infiltration of ocular nerves or muscles. Visual loss secondary to optic nerve involvement is usually a late sign, although more subtle signs of optic nerve dysfunction, including afferent pupillary defect, loss of color vision, and visual field defect are more frequently encountered. Finally, orbital involvement may be asymptomatic, and is only discovered on CT or MRI evaluation of patients with sinonasal complaints. Evaluation of patients with suspected primary or secondary malignancy in the orbit should include a detailed neuro-ophthalmologic examination. This usually includes detailed assessment of visual acuity, visual fields, and ocular motility. Other ophthalmologic evaluation includes careful pupillary examination for afferent pupillary defect or anisocoria, external examination to include Hertel exophthalmometry, and marginal reflex distance as an indicator of eyelid position. Slit lamp examination of the conjunctivae, cornea, anterior chamber, and lens is appropriate. Finally, detailed examination of the fundus may reveal compressive effect, intraocular malignancy, or an unrelated reason for visual loss. Formal testing of color vision and automated visual fields is commonly appropriate.
IMAGING Indications Imaging of the nasal cavity, paranasal sinuses, and orbit is indicated whenever there is clinical suspicion of a neoplastic process. Imaging is also indicated for obtaining pretreatment information regarding the location, size, extent, and invasiveness of the primary tumor, as well as the presence of regional and distant metastasis. Such information is crucial in deciding on therapeutic options and for proper preoperative planning of the optimal surgical approach. Imaging also plays an important role in the posttreatment follow-up, indicating areas of residual or recurrent disease, and defining suspicious areas for biopsy.
238
Hanna et al.
Ethnic distribution
Gender distribution 2500
77.7%
2000 Number of patients
Number of patients
2000 1500 60.7%
1000
39.3%
500
1500 1000 9.1%
500
0
Male
White
Number of Patients
Age distribution
1.8%
Black
Hispanic Ethnicity
Asian
Other
Histology
34.2% 28.2%
Malignant melanoma 6%
21.4%
Neuroblastom a 4%
Other neoplasms 5% Squamous carcinoma 45%
Sarcoma 10%
8.1% 4.3%
3.7%
< 20
0.3%
0
Female Gender
1000 900 800 700 600 500 400 300 200 100 0
11.1%
20 – 35 35 – 50 50 –65 65 – 80
80 +
Age group
Other carcinoma 7% Adenocarcinoma 6%
Unclassified carcinoma 9%
Adenoid cystic carcinoma 8%
Site distribution
Number of patients
1400
10%
44.9%
1200 1000
12% 34.3%
800 600 400 200 0
11.3%
6.3%
2.0%
1.3%
58%
20%
Not Maxillary Nasal Ethmoid Sphenoid Frontal specified sinus cavity sinus sinus sinus
Site
Figure 12 Patients with sinonasal cancer seen at MD Anderson Cancer Center between 1944 and April 2007. (N = 2698 patients). Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
Imaging Modalities Both computed tomography (CT) and magnetic resonance imaging (MRI) may be needed for optimum radiologic assessment of sinonasal malignancy, particularly in assessing the cranial base, orbit and pterygo-palatine, and infratemporal fossae. Coronal images best delineate involvement of the orbital walls and invasion of the skull base, particularly the cribriform plate. Axial images are particularly helpful in demonstrating tumor extension through the posterior wall of the maxillary sinus into the pterygopalatine fossa and infratemporal fossae. Sagittal images are particularly helpful in evaluating extension along the cribriform plate, planum sphenoidale, and clivus (Fig. 16). The main advantage of CT scans is in delineating the architecture of the bones, especially in “bone windows” [Fig. 16(A) and 16(B)]. The addition of contrast enhancement
increases tumor definition from adjacent soft tissue, especially intracranially. Bone destruction and soft tissue invasion suggest an aggressive lesion, usually a malignant neoplasm. Widening or sclerosis of the foramina of the infraorbital, vidian, mandibular, or maxillary nerves may indicate perineural spread (Fig. 17). MRI is unsurpassed in delineating soft tissue detail, both intra- and extracranially (Fig. 16). Obliteration of fat planes in the pterygopalatine fossa, infratemporal fossa, and nasopharynx usually indicates tumor transgression along these boundaries. Dural thickening or enhancement is usually an indication of tumor involvement, and evaluation of critical structures such as the brain and carotid artery is best delineated by MRI. Similarly, enhancement or thickening of cranial nerves indicates perineural spread, which is better detected on MRI than CT (12) (Fig. 17). Perhaps one of the most
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
239
significant advantages of MRI is the ability to distinguish tumor from retained secretions secondary to obstruction of sinus drainage (Fig. 18). MRI is also particularly helpful in monitoring patients in the postoperative follow-up period, although this role may be supplanted in the near future by positron emission tomography (PET) scans because of its ability to distinguish between tumor recurrence and posttreatment fibrosis. PET–CT is also helpful in delineating regional and distant metastasis (Fig. 19). Angiography is not indicated in the routine assessment of patients with neoplasms of the nose, paranasal sinuses, and orbit. In certain selected cases, however, angiography may be necessary. These cases include vascular neoplasms of the sinonasal region, where angiography will not only delineate the tumor extent and the blood supply but also permit the use of selective embolization of the vascular supply to the tumor (Fig. 20). This reduces intraoperative blood loss, facilitating surgical resection. Figure 13 Carcinoma of the nasal cavity. Endoscopic view showing a tumor arising from the floor of the right nasal cavity. Biopsy revealed squamous cell carcinoma. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
(A)
(B)
Figure 14 Advanced ethmoid carcinoma. Clinical photographs showing the mass centered on the nasion, and causing widening of the interpalpebral distance (telecanthus). The mass shows involvement of the overlying skin and destruction of the underlying nasal bone. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
BIOPSY Nasal Cavity and Paranasal Sinuses The definitive diagnosis of a neoplasm of the nasal cavity and paranasal sinuses relies on expert histopathologic review of any biopsy specimens by a head and neck pathologist to confirm the exact diagnosis prior to treatment. This is critical since the treatment and prognosis of sinonasal cancer is greatly influenced by histology (13). The vast majority of sinonasal neoplasms are accessible for biopsy through a strictly endonasal approach. A wide variety of rigid nasal endoscopes offers superb visualization of intranasal lesions with a high degree of optical resolution and bright illumination (Fig. 13). The application of topical anesthetics and decongestants improves visualization and allows thorough examination of the nasal cavity. The site of origin of the lesion and its relation to the nasal walls (septum, floor, roof, and lateral nasal wall) should be noted. An adequate specimen should be obtained, avoiding crushing of tissue, and submitted for histopathologic examination. If the diagnosis of lymphoma is suspected, fresh tissue should be sent in saline, rather than fixed in formalin. Most endonasal biopsies can be performed in the outpatient setting with minimal discomfort to the patient. In certain cases, the diagnosis of a highly vascular neoplasm, such as angiofibroma, may be suspected on clinical grounds. Under these circumstances, it is prudent not to perform the biopsy until imaging and angiography (possibly with embolization) are performed (Fig. 20). Preoperative biopsy can then be performed in the operating room under controlled conditions to confirm the diagnosis before surgical resection. If a nasal mass is suspected to have an intracranial communication such as an encephalocele, meningocele, or nasal glioma, this should be confirmed with imaging to avoid inadvertent CSF leak and subsequent meningitis (Fig. 21).
Orbit (A)
(B)
Figure 15 Carcinoma of the maxillary sinus may extend inferiorly and destroys the palate presenting as an ulcer (A) or submucosal mass (B). Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
In most cases of primary intraorbital tumors, the approach used to obtain a biopsy is dictated by the location of the tumor. Lesions in the superior orbit may be addressed by a coronal flap or through a brow incision (modified Kronlein or Stallard-Wright incision). Lateral orbital lesions may require removal of the orbital rim. The transconjunctival approach, with detachment of the lateral canthal tendon, provides access to the orbital floor. The medial orbit may be entered
240
Hanna et al.
(A)
(C)
(B)
(D)
Figure 16 Ethmoid carcinoma. Coronal and sagittal images showing bony destruction on CT scan (A and B) and intracranial invasion on T1-MRI with contrast (C and D). Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
through a modified Lynch or transcaruncular incision. Lesions within the muscle cone may be addressed by elevating conjunctiva and Tenon’s capsule from the globe and detaching the necessary rectus muscle. Lacrimal gland lesions are best approached through the upper lid crease (1).
PREOPERATIVE PREPARATION A thorough preoperative assessment should determine the candidacy of a patient for surgical management of his or her neoplasm. This involves a careful “mapping” of the tumor extent, as well as the general medical condition and functional status of the patient. This is usually accomplished by a detailed history and physical examination, and comprehensive examination of the head and neck region including endoscopy of the sinonasal region. Cranial nerve examination as well as ophthalmologic evaluation should
be done to evaluate cranial base and orbital extension, respectively. High-resolution imaging should be obtained using CT or MRI, or both, to accurately assess the tumor extent. In certain cases, angiography will be needed to determine the extent of carotid arterial involvement. The balloon occlusion test should be performed if carotid artery resection or reconstruction is contemplated. Preoperative embolization may be indicated in certain vascular tumors. Neurosurgical consultation is needed if a combined craniofacial approach is anticipated. If free vascularized flaps will be used for reconstruction, expertise with microvascular surgery is needed, and appropriate consultation should be obtained. Evaluation by a maxillofacial prosthodontist is required in most patients to obtain preoperative dental impressions and design surgical obturators or splints for maintenance of proper dental occlusion and oral rehabilitation. Similar expertise is essential in cases where prosthetic
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
(A)
241
(B)
Figure 17 Perineural spread of adenoid cystic carcinoma along the third division of the trigeminal nerve: (A) A coronal CT with IV contrast showing widening of the left foramen ovale (black arrow), compared to the one on the right. There is also an enhancement and thickening along the left Meckel cave (white arrows). (B) A coronal T1-weighted MRI with gadolinium showing marked thickening and enhancement of V3, trigeminal ganglion, and the lateral cavernous sinus (CS). The tumor abuts the cavernous carotid artery (white arrow). There is enhancement of the dura along the floor of the middle cranial fossa (black arrow). This “dural tail” is usually a sign of involvement of the dura with tumor. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
(A)
(B)
Figure 18 Sinonasal Melanoma. (A) Coronal CT scan demonstrating opacification of the right nasal cavity as well as the maxillary and ethmoid sinuses. There appears to be destruction of the lateral nasal wall and the nasal septum. The lesion is abutting the orbital floor and the cribriform plate, but it is unclear whether or not these structures are involved. (B) Coronal T1-weighted MRI with gadolinium of the same patient revealing that the lesion is limited to the nasal cavity and ethmoid sinuses and that the changes in the maxillary sinuses are due to retained secretions secondary to obstruction of the ostium, rather than soft tissue involvement. It also demonstrates that the lesion does not invade the orbit or the cranial base. The presence of low signal areas within the lesion gives it a heterogeneous appearance, which is characteristic of sinonasal melanoma. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
242
(A)
Hanna et al.
(B)
(C)
Figure 19 PET–CT of head and neck. These images are from the same patient whose CT and MRI are depicted in Figure 16. The fused PET–CT images show ethmoid carcinoma (A) with metastasis to the retropharyngeal lymph nodes (B, C), which were not detected on CT or MRI. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
orbital, nasal, or facial rehabilitation is required. Consultations with ophthalmology should be considered for detailed neuro-ophthalmologic examination of all patients with known or suspected orbital involvement. If orbital exenteration is contemplated, visual function of the contralateral eye should be carefully assessed. Consultations with medical and radiation oncology colleagues should be done to consider incorporation of chemotherapy or radiation in the treatment plan. Radiation and/or chemotherapy may be used preoperatively as induction (neoadjuvant), or postoperatively as adjuvant therapy. This is particularly important in patients with advanced stage disease (e.g., dural or orbital involvement) or high-grade lesions (e.g., sinonasal undifferentiated carcinoma. In selected cases, chemotherapy and/or radiation may be reasonable alternatives to surgery. Such decisions are best discussed in the format of a multidisciplinary tumor board. If surgery is chosen as a treatment modality, the plan for the surgical approach, the extent of resection, and reconstructive options should then be formulated. This plan should be communicated clearly among the various members of the surgical team, particularly the otolaryngologists, head and neck surgeons, neurosurgeons, and plastic and reconstructive surgeons. Careful assessment of the patient’s general medical condition should be carried out prior to surgery. Preoperative chest radiograph, blood counts, liver and renal function tests, blood sugar, electrolytes, coagulation studies, and an electrocardiogram should be performed routinely. Appropriate consultations from medical colleagues should be obtained in order to optimize the patient’s medical status before surgery, and help in management postoperatively. The patient’s nutritional status should be evaluated, and if indicated, enteral or parenteral feeding may be considered. High-resolution imaging for metastatic workup is not routinely performed, unless indicated by history, clinical examination, chest radiograph results, or blood test abnormalities. Finally, the surgical team should discuss with the patient and family the nature of the disease, the evaluation, the indications, risks, possible complications, sequelae, and
alternatives of therapy. The expected postoperative course including length of stay in the hospital, feeding, rehabilitation, and need for adjunctive therapy should be described. This ongoing communication should be maintained in a clear, honest, and sympathetic fashion throughout the course of patient care.
SURGERY Indications and Contraindications Surgery is indicated when there is adequate evidence that the tumor can be completely resected with acceptable morbidity. For early stage disease (T1–T2), surgery alone may be adequate treatment, but for more advanced stage resectable disease, postoperative adjuvant radiation or chemoradiation is commonly used to improve tumor control (14). In the presence of tumor extension to the cavernous sinus, internal carotid artery, optic chiasm, extensive brain parenchymal involvement, or distant metastasis, surgery is usually contraindicated. However, in selected cases, surgery with proper adjuvant therapy may still offer the most effective local disease palliation even in the presence of extensive disease.
Surgical Principles When dealing with the subject of surgical treatment of sinonasal cancer, a distinction has to be made between the terminology used to describe the surgical approach on the one hand, and the extent of resection on the other hand. A surgical approach describes the various incisions, soft tissue dissection, and skeletal osteotomies required to expose the tumor and adjacent structures to perform a complete and safe resection. On the other hand, the extent of tumor resection describes the various structures that need to be surgically extirpated to achieve total tumor removal with tumor-free margins. Obviously, both the surgical approach and the extent of resection are closely related and depend on the extent of tumor, its aggressiveness, and related critical structures. The various surgical approaches and extent of resection are listed in Table 2 (1). The choice of surgical approach and
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
(A)
(B)
(C)
(D)
243
Figure 20 Juvenile Nasopharyngeal Angiofibroma.Coronal (A, B) and axial (C) CT with contrast and axial T1-axial MRI with contrast (D) showing the mass in the nasal cavity, maxillary sinus, nasopharynx, sphenoid sinus, pterygoid plates, and pterygo-palatine and infratemporal fossa. The mass involves the floor of the middle cranial fossa and extends intracranially to the cavernous sinus. Continued
extent of resection will generally depend on the location and the extent of the tumor. In some cases, different approaches may be equally effective for resection of a particular tumor. For example, a tumor of the nasal cavity, lateral nasal wall, ethmoid, sphenoid and medial maxillary sinus requiring a medial maxillectomy and a total sphenoethmoidectomy may be adequately resected using a transfacial, endoscopic, or sublabial approach (Figs. 18 and 22). However, the following principles should always guide the surgeon in
choosing the optimal approach and extent of resection for all patients undergoing surgical treatment of sinonasal cancer:
r r r r r
Adequate oncologic resection Minimal brain retraction Protection of critical neurovascular structures Meticulous reconstruction of the skull base Optimal esthetic outcome
244
Hanna et al.
(E)
(F)
(G)
(H)
Figure 20 T1-sagittal MRI (E, F) shows flow voids of increased extensive vascular supply coming from the internal maxillary artery (E) and internal carotid artery (F). Early phase angiogram showing the blood supply from the internal maxillary artery (G) and late phase showing significant tumor vascular blush and contribution from the internal carotid artery (H). Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission. Continued
Surgical Approach Endonasal Approaches Endoscopic endonasal approaches (EEA) are being increasingly used for surgical excision of selected tumors of the sinonasal tract, either alone or in combination with open approaches. The long-term oncologic results of EEA in treating malignant sinonasal tumors are still being defined. The keys to adequate oncologic results using EEA are good selection of patients and the surgeon’s experience with this approach. EEA are most suited for central tumors involving the nasal septum, nasal cavity and lateral nasal wall, ethmoid and sphenoid sinuses, and clivus (Fig. 22). The advantages of EEA include direct bilateral access to the tumor, superior illumination, magnification and visualization of the surgical field (Fig. 7), wider angles of vision using angled endoscopes, and relatively less morbidity compared to the open surgical approaches. The limitations of EEA include lack of binocular vi-
sion and depth perception, which is important when dealing with critical neurovascular structures. Ergonomic limitations include the inability of the primary surgeon to control the endoscope and two instruments simultaneously and hence the reliance on a surgical assistant for camera control during twohanded surgery. The major limitation of EEA is the inability to repair or patch dural defects with suture techniques, which limits the reconstructive options after endoscopic resection of intradural tumors. We are currently exploring new applications in robotic surgery to overcome some of these limitations (15) (Fig. 23).
Transfacial Approaches: Lateral Rhinotomy and Weber-Fergusson Transfacial approaches are the most commonly used surgical approaches for resection of locally advanced sinonasal
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
(A)
245
(B)
Figure 21 Meningo-encephalocele. Sagittal (A) and coronal (B) T1-MRI with gadolinium showing a defect and in the anterior cranial base at the right fovea ethmoidalis and cribriform plate. There is herniation of a meningo-encephalocele, which presented as a nasal mass. Inadvertent biopsy of such lesions may lead to CSF leaks and should be avoided. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
tumors. They allow adequate exposure of the nasal cavity, maxillary sinus, pterygo-palatine fossa, pterygoid plates, ethmoid sinuses, medial and inferior orbital walls, sphenoid sinus, nasopharynx, clivus, and the medial aspect of the infratemporal fossa. The lateral rhinotomy is the standard incision for exposure of sinonasal tumors through a transfacial approach (Fig. 24). It can be used alone, or various extensions of the basic incision may be added for further exposure depending Table 2 Surgical Treatment of Sinonasal Cancers Surgical approach Endoscopic Lateral rhinotomy and Weber-Fergusson Transoral–Transpalatal Facial “Degloving” Craniofacial Extent of resection Ethmoidectomy Inferior maxillectomy Medial axillecotmy Total maxillectomy Anterior cranial base resection Infratemporal fossa dissection Orbital exenteration
on the extent of tumor (1). The Weber-Fergusson incision adds a lip-splitting and subciliary incision for added exposure of the maxillary bone. We prefer to extend the lateral rhinotomy toward the medial brow, a Lynch-type extension, and avoid the subciliary incision of the classic Weber-Fergusson to minimize eyelid complications, as will be discussed further under the section on “Total maxillectomy”. The basic lateral rhinotomy incision provides adequate exposure when performing a medial maxillectomy. Elevation of the soft tissues of the cheek is done in a subperiosteal plane over the maxilla and around the inferior orbital nerve [Fig. 25(A)]. The attachment of the medial canthal tendon to the nasal bone is released. The periorbita is elevated over the medial orbital wall exposing the lacrimal crest, the lamina papyracea, and the frontoethmoidal suture. This suture serves as a landmark for the position of the floor of the anterior cranial fossa, and when followed posteriorly, leads to the anterior and posterior ethmoidal foramina. The anterior and posterior ethmoidal arteries are cauterized with the bipolar electrocautery, clipped or ligated, and transected [Fig. 25(B)]. The optic nerve is located 4- to 5-mm posterior to the posterior ethmoidal artery. The orbital floor should be dissected as far lateral as the inferior orbital fissure. The lacrimal sac is identified in its fossa between the anterior and posterior lacrimal crests. If a medial maxillectomy is performed, the lacrimal sac is elevated from the fossa, and the lacrimal duct
246
(A)
Hanna et al.
(B)
Figure 22 Hemangiopericytoma of the sinonasal region. Preoperative (A) and 3 years postoperative (B) MRI of a patient who underwent endoscopic resection of hemangiopericytoma of the sinonasal tract. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
(A)
(C)
(B)
(D)
Figure 23 Robotic assisted endoscopic surgery of the anterior skull base (Cadaveric Dissection). (A) Soft tissue approach: Bilateral sublabial incisions and soft tissue flap elevation. (B) Bilateral anterior maxillary antrostomies. (C) Robotic ports placement. The camera port is placed into the right nostril, and the right and left surgical arm ports through the respective anterior then middle antrostomies. (D) Bimanual sharp dissection of the mucosa covering the fovea ethmoidalis and cribriform plate.
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
(E)
247
(F)
(G)
Figure 23 (E) Wide sphenoidotomy with excellent access to the sella turcica (ST) and para-sellar region (PS). (F) The cribriform plate (CP) is removed bilaterally, and the cut edges of the olfactory nerves (ON) are shown. The dura is incised or resected to expose the inferior surface of the frontal lobes (FL) intracranially. (G) Dural repair. Suturing the dural edges, making a loop, and tightening the knot. Source: From Ref. 15.
transected, and the sac marsupialized into the nasal cavity to provide adequate drainage of the lacrimal system and to prevent stenosis [Fig. 25(C)].
Midfacial Degloving and Sublabial Approaches The midfacial degloving approach is most commonly used in the management of large benign lesions of the sinonasal region and the skull base, such as juvenile nasopharyngeal angiofibroma, for selected malignancy in this area and to afford access to the nasopharynx and infratemporal fossa. The main advantage of the “degloving” approach is that an external facial incision is avoided. Another advantage is providing simultaneous exposure to the inferior and medial maxilla, bilaterally (Fig. 26). This is particularly helpful when approaching tumors with bilateral involvement of the nasal cavity and maxillary sinus. A major disadvantage, however, is the limited superior and posterior exposure, and the need for constant retraction of the soft tissue envelope for continued adequate exposure. The midfacial degloving approach requires a basic level of proficiency and understanding of closed rhinoplasty incisions. It involves a complete transfixion incision of the membranous septum. This is joined endonasally with a bilateral
intercartilaginous incision, with soft tissue elevation over the nasal dorsum as far superior as the nasal root. The nasal skeleton is therefore “degloved” from overlying soft tissues as far lateral as the pyriform aperture. A gingivobuccal incision extends bilaterally across the midline to both maxillary tuberosities laterally. Subperiosteal dissection is continued cephalad over the face of both maxillae. The dissection joins the nasal degloving using sharp dissection over the pyriform aperture attachments (Fig. 26). The sublabial approach may also be used to access tumors of the sphenoclival region such as chordoma, particularly if the lesion extends lower than the horizontal plane of the palate, for example the lower third of the clivus and craniovertebral junction (Fig. 27). A Le Fort I osteotomy is done and the maxilla is displaced inferiorly after posterior osteotomies separate the maxilla from the pterygoid plates. We prefer a combination of unilateral or half a Le Fort I osteotomy with a median or paramedian palatal osteotomy for better displacement of the maxilla inferiorly and laterally. This offers a wider exposure because it avoids the cantilever effect of the posterior maxilla upward restricting exposure when the anterior maxillary segment is displaced inferiorly when the bilateral (complete) Le Fort osteotomy is used.
248
Hanna et al.
below the level of the frontoethmoidal suture to avoid injury to the floor of the anterior cranial fossa. The superior attachment of the middle turbinate is then transected along the roof of the nose. Posteriorly, the lateral nasal wall cuts are connected with right-angled scissors behind the turbinates. The specimen is thus delivered and examined for margins with frozen section control. If the tumor involves the nasal septum, it should be included in the resection specimen by adding appropriate septal cuts to allow for tumor-free margins. Closure begins by reattachment of the medial canthal tendon to the nasal bone in its anatomic position. Meticulous multilayered closure of the lateral rhinotomy is performed and usually results in excellent healing and acceptable postoperative appearance [Fig. 25(E)]. If a sublabial approach is done, the mucosal incisions are closed with absorbable suture. Nonadherent nasal packing may be left for 1 to 2 days.
Inferior Maxillectomy Figure 24 Lateral rhinotomy incision (1, 2, and 3) and its extensions (4 and 5). A temporary tarsorrhaphy protects the ipsilateral globe. The basic lateral rhinotomy incision is outlined by connecting three surface points. The first point (1) is marked half way between the nasion (A) and the medial canthus (B). The second point (2) is where the alar crease begins and the third point (3) is at the base of the columella. The basic incision provides adequate exposure for a medial maxillectomy. The basic incision may be extended to include a lip-splitting extension (4) or a “Lynch” type extension (5) if further exposure is necessary. The extended incision provides adequate exposure for a total maxillectomy. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
This procedure involves resection of the inferior maxillary sinus below the plane of the infraorbital nerve. It is most commonly used for neoplasms of the alveolar process of the maxilla with minimal extension to the maxillary antrum. Similarly, lesions of the hard palate sparing the antrum can be treated by an inferior maxillectomy. A combination of sublabial and palatal incisions is usually used for exposure and osteotomies are done around the lesion, ensuring an adequate margin of resection (Fig. 28). Alternatively, a midfacial degloving can be used for lesions crossing the midline and involving the inferior maxilla bilaterally.
Total Maxillectomy
Extent of Resection Medial Maxillectomy The most common indication for medial maxillectomy is in the treatment of tumors of the nasal cavity, lateral nasal wall, and medial maxillary sinus (Fig. 18). Medial maxillectomy includes removal of the lateral nasal wall, and the medial maxillary segment bounded laterally by the infraorbital nerve. In addition, a complete sphenoethmoidectomy is usually performed. The incision most commonly used for exposure is the lateral rhinotomy (Fig. 24). Alternatively, a midfacial degloving, as will be described later, may be used, and is preferable if bilateral medial maxillectomy is needed (Fig. 26). Endonasal endoscopic medial maxillectomy and sphenoethmoidectomy may also be performed for appropriately selected tumors (Fig. 22). If the lateral rhinotomy is performed and soft tissue exposure is completed as discussed in the previous section, osteotomies are done as shown in Figure 25(D), and the anterior wall of the maxillary sinus above the level of dental roots and medial to the infraorbital nerve is removed. Lateral to the infraorbital foramen, the anterior wall antrostomy may be enlarged to expose the zygomatic recess of the antrum. Resection of the lateral nasal wall begins with an inferior osteotomy along the nasal floor below the attachment of the inferior turbinate, starting at the pyriform aperture, and carried posteriorly to the posterior maxillary wall. With the orbit retracted laterally and protected with a malleable brain retractor, the lamina papyracea is identified and, if necessary, resected. A complete sphenoethmoidectomy is done, staying
If the extent of resection requires a total maxillectomy, the lateral rhinotomy incision may be extended by adding lipsplitting, gingivo-buccal, and palatal incisions inferiorly. The lip-splitting incision, which may be done along the philtrum or in the midline, connects the lateral rhinotomy with the sublabial incision, thus allowing more lateral elevation of the facial flap. The gingivo-buccal incision starts from the lip-splitting incision and extends as far laterally as the region of the first molar and over the lateral surface of the maxillary tuberosity. In patients undergoing total maxillectomy, a median or paramedian palatal incision is performed over the hard palate extending from an inter-incisor space anteriorly, to the junction of the soft and hard palate posteriorly. The incision then continues laterally between the hard and the soft palate to curve posterolaterally around the maxillary tuberosity meeting the gingivo-buccal incision [Fig. 29(A)]. In patients undergoing total maxillectomy with orbital preservation, we prefer to extend the lateral rhinotomy superiorly beneath the medial brow rather than laterally through a subciliary incision used in the classic Weber-Fergusson approach [Fig. 29(B)]. We described several advantages to this modification (16). First, avoiding a subciliary incision eliminates any disruption to the lower lid skin–muscle– tarsus complexthat minimizes lower eyelid complications, particularly ectropion and prolonged eyelid edema. Another advantage is avoiding trifurcation of the incision reducing the risk of skin breakdown at the medial canthal area. This is especially important for previously irradiated patients, who are more prone to develop medial canthal skin dehiscence. Similarly, since the vascularity of the thin lower eyelid skin is not affected with the extended lateral rhinotomy incision, patients who undergo orbital floor reconstruction with implants such as titanium mesh have less chance to
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
249
(A)
(B)
(C)
(D)
(E)
Figure 25 Lateral rhinotomy and Medial Maxillectomy. (A) Elevation of the soft tissues of the cheek is done in a subperiosteal plane over the maxilla and around the inferior orbital nerve (ION). The periorbita is elevated over the anterior lacrimal crest (ALC) to expose the lacrimal sac (LS). (B) Dissection of the medial periorbita over the lamina papyracea reveals the anterior ethmoid artery (arrow) at the level of the frontoethmoidal suture line, which marks the level of the anterior cranial floor. The artery is coagulated by bipolar electrocautery, clipped or ligated and then transected. (C) After the lacrimal sac is transected, it is marsupialized into the surgical cavity as a dacryocystorhinostomy. Silicone stents are placed through the upper and lower canaliculi and brought into the nasal cavity to prevent postoperative epiphora. These stents are removed in about 3 to 6 months. (D) Medial Maxillectomy. Osteotomies: (A) vertically medial to the infraorbital foramen (arrowhead), (B) horizontally above the level of dental roots and into the pyriform aperture, and (C) obliquely along the nasomaxillary suture line. If the lateral nasal wall is to be resected, the lacrimal sac (arrow) is transected and marsupialized into the nasal cavity. (E) Postoperative appearance of a lateral rhinotomy incision. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
250
Hanna et al.
Figure 26 Sublabial “facial degloving” approach. In addition to avoiding facial incisions, the sublabial approach has the advantage of providing bilateral access to the medial and inferior maxillary segments. Source: This illustration in the property of The Department of Neurosurgery, M.D. Anderson Cancer Center is used with permission.
develop wound breakdown and implant exposure. While the extended lateral rhinotomy incision has several functional and cosmetic advantages, it does not compromise exposure and provides an adequate approach for a safe oncologic resection. The extension of the lateral rhinotomy incision beneath the medial eyebrow shifts the fulcrum of rotation of the soft tissue flap superiorly and laterally enhancing lateral exposure, which is not less from that obtained with a classic Weber-Fergusson incision [Fig. 29(B)]. Transection of the infraorbital nerve allows even more lateral and posterior elevation of the soft tissues, to expose the entire maxillary bone as far lateral as its zygomatic extension, and posteriorly to the pterygomaxillary fissure and over the pterygoid plates. Additionally, its postoperative cosmetic appearance is superior to the Weber-Fergusson incision [Fig. 25(E)]. Whichever incision is used, elevation of the facial flap is usually done in the subperiosteal plane. However, if the tumor has invaded the anterior wall of the maxillary antrum, a supraperiosteal plane is used. Occasionally, the cheek skin overlying the maxilla is included with the specimen, if it is involved with tumor. With the globe protected with a temporary tarsorrhaphy stitch, the periorbita is dissected along the medial, inferior, and lateral orbital walls.
Lateral osteotomies are performed along the frontal and temporal processes of the zygoma [Fig. 29(B)]. Medial osteotomies are done along the frontal process of the maxilla and along the medial orbital wall just below the frontoethmoidal suture, extending posteriorly to the level of the posterior ethmoidal foramen. The medial and lateral osteotomies are then connected superiorly across the orbital floor along the inferior orbital fissure. Inferiorly, a midline sagittal osteotomy is made across the hard palate. The ipsilateral central incisor should be preserved, if possible, to enhance prosthesis retention. Finally, after the internal maxillary artery is identified at its entrance through the pterygomaxillary fissure, ligated, and transected, a posterior osteotomy is done to disarticulate the maxilla from the pterygoid plates. The maxilla is delivered by anteroinferior traction, while remaining soft tissue attachments are cut using a curved heavy scissors. Bleeding is usually encountered at this point, and is controlled by temporary packing of the cavity, followed by electro coagulation of bleeding mucosal surfaces or ligature of bleeding points. The pterygoid plexus of veins may be a source of persistent bleeding, and can be managed by hemostatic figure-of-8 sutures and surgical packing. Bleeding is usually minimized if the internal maxillary artery is ligated before the posterior osteotomy is done along the pterygomaxillary fissure. Total maxillectomy usually involves removal of the entire maxillary bone including the palate and the orbital floor [Fig. 29(C)]. Preservation of the orbital floor (subtotal maxillectomy) or the palate (suprastructure maxillectomy) may be done if these structures are not involved by tumor. Depending on the extent of the lesion, resection may extend beyond the posterior wall to the pterygo-palatine fossa and pterygoid plates. Perineural spread of tumor along V2 may be resected by following the nerve through foramen rotundum and into Meckel cave (17) (Fig. 30).
Craniofacial Resection Surgical resection of the anterior cranial base is commonly indicated for patients with sinonasal tumors involving the cribriform plate or fovea ethmoidalis. This is done, by definition, for most cases of esthesioneuroblastoma, as well as carcinoma of the ethmoid or maxillary sinuses approaching or involving the anterior cranial base (Fig. 16). Tumors transgress the cribriform plate either by direct bony invasion or by perineural spread along the filaments of the olfactory nerves. The dura of the anterior cranial fossa forms a barrier that delays, to a certain extent, brain invasion. Dural resection in patients with intracranial but extradural disease or patients with limited dural involvement often provides an adequate oncologic margin. However, malignant tumors that transgress the dural barrier and involve the underlying brain parenchyma are usually associated with poor prognosis (18). However, even in some cases with limited frontal lobe involvement, anterior craniofacial resection may still be indicated for local control of the disease. Resection of the floor of the middle cranial fossa is sometimes performed in patients with sinonasal tumors to achieve tumor-free surgical margins for lesions extending to the roof of the infratemporal fossa or for those tumors that exhibit perineural spread along the branches of the trigeminal nerve to the Gasserian ganglion, most commonly adenoid cystic carcinoma. Craniofacial approaches combine extra- and intracranial access to the anterior and lateral skull base. Extracranial
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
(A)
(C)
251
(B)
(D)
Figure 27 Sublabial approach and inferior maxillotomy for access to sphenoclival region. (A) Preoperative (upper) and postoperative (lower) coronal and sagittal MRI of a patient with clival chordoma. (B) Le Forte I Osteotomy (black arrow) is done on one side only and connects the pyriform aperture. A second paramedian palatal osteotomy is performed (white arrow). (C) Inferior displacement of the maxilla (maxillotomy) to expose the sphenoid sinus, nasopharynx, and clivus. (D) Rigid fixation of the maxillary segments using “preregistered” titanium microplates to avoid any malocclusion. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
approaches may include transfacial, sublabial, or endonasal approaches as previously described (Fig. 31). The bicoronal incision starts in a preauricular crease anterior to the tragus. The superficial temporal artery should be dissected and preserved. The scalp incision is extended in
the coronal plane, staying behind the hairline along its entire course, to the contralateral preauricular region. We prefer to gently curve the incision anteriorly at the midline [Fig. 32(A) and 32(B)]. The scalp flap is elevated in a subgaleal plane superficial to the pericranium between the superotemporal
252
Hanna et al.
(A)
(B)
Figure 28 Inferior maxillectomy. Gingivo-buccal and palatal incisions and osteotomies are done around the lesion as shown in (A). If there is adequate space between the central incisors, the osteotomy may be performed using a micro–reciprocating saw in the interincisor space. Otherwise, the ipsilateral central incisor should be extracted and the osteotomy placed in the tooth socket in order to avoid loss of bony support to the remaining contralateral incisor. Inferior maxillectomy specimen (B) showing adequate surgical margins around an upper alveolar ridge carcinoma (arrow). Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
lines bilaterally. Lateral and inferior to the superotemporal lines, an incision is made through both the superficial and deep layers of the temporalis fascia 1 to 1.5 cm posterior to the superior orbital rim and extends posteriorly parallel to the course of the zygomatic arch [Fig. 32(C)]. Dissection proceeds below the plane of the deep layer of the temporalis fascia to preserve the frontal branch of the facial nerve, which is superficial to the fascia. The scalp flap is elevated anteriorly toward the superior orbital rims and posteriorly toward the vertex. Pericranial incisions are made as far posteriorly as necessary to provide adequate length for the pericranial flap and along the superotemporal lines bilaterally. The pericranial flap is dissected free from the underlying bone and reflected anteriorly [Fig. 32(D)]. Careful dissection and preservation of the supraorbital neurovascular pedicles is necessary to provide a well-vascularized pericranial flap for reconstruction of the cranial base defect. The supraorbital nerves and vessels are located along the medial one-third of the superior orbital rim. Elevation of the supraorbital rim periosteum begins laterally and proceeds medially, until the margin of the supraorbital groove is carefully exposed with a fine elevator. The nerve and vessels may exit the skull either through a notch or a true foramen. If a notch is present then the nerve can be dissected free without difficulty. If a foramen rather than a notch is found, the floor of the foramen is removed with a fine osteotome. This liberates the pedicle and further elevation of the superior periorbita is then achieved. Frontal, temporal, or frontotemporal craniotomy is then performed to allow access to the floor of the anterior or middle cranial fossa or both, respectively. For a frontal craniotomy, bilateral burr holes are then placed in the depression posterior to the frontal-zygomatic sutures (MacCarty keyhole), after reflection of the temporalis muscle leaving a cuff of fascia at the superotemporal line for reattaching the muscle during closure [Fig. 32(E)]. These anatomic keyholes provide access to the anterior fossa dura and by inferior enlargement
the periorbita, if needed. A burr hole is then placed on the superior sagittal sinus, well anterior to the coronal suture, exposing the dura on both sides of the sinus. The bifrontal craniotomy is performed between the burr holes. The craniotomy may be extended inferiorly to the level of the frontonasal suture. Compared to frontal craniotomy, subfrontal approaches have the advantage of minimizing brain retraction by providing wider and more direct exposure of the floor of the anterior cranial fossa (Fig. 33). This is especially helpful in more posteriorly located lesions such as those involving the planum sphenoidale, clivus, orbital apex, and optic chiasm (Fig. 34). The subfrontal approach is done by adding osteotomies that allow incorporation of the superior orbit and/or nasal bone to the craniotomy. These skeletal elements may be removed in several subunits or as a single bone flap (10) (Fig. 35). Bilateral nasal osteotomies are done along the lower border of the nasal bones and then along the suture line between the nasal and lacrimal bones [Fig. 35(C)]. The osteotomies are connected across the midline below the frontoethmoidal suture line and in front of the anterior ethmoidal vessels. This avoids injury to the cribriform plate and olfactory nerves. After the dura and periorbita have been carefully dissected from the bone, the lateral wall and roof of each orbit are removed in separate orbital osteotomies [Fig. 35(C)]. Under direct visualization, taking care to protect the periorbita and dura, an anteroposterior cut is made at the medial aspect of the orbital roof staying lateral to the ethmoid sinus. A second anteroposterior cut is made at the inferior aspect of the lateral orbital wall. These cuts are connected posteriorly with protection of the tissues of the SOF. The bone flap consisting of the frontal bone, orbital roof, superolateral orbital rims, and nasal bones can be removed for wide exposure [Fig. 35(D) and 35(E)]. After completing the craniotomy, brain “relaxation” is achieved by opening the dura and allowing release of CSF or by withdrawing CSF from a lumbar subarachnoid drain,
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
253
(A)
(B)
(C)
(D)
Figure 29 Total maxillectomy. (A) Intraoral incisions. (B) The exposure offered through an extended lateral rhinotomy shown in this figure is not less than that offered by the Weber-Fergusson incision. The advantage of the former is avoiding a subciliary incision and its potential for lower eyelid ectropion and edema. Osteotomies have been performed as indicated by the arrows.(C) En bloc resection specimen. Note tumor involvement of the orbital floor, which had to be resected. The tumor did not transgress the periorbita, therefore the eye was preserved. (D) Total maxillectomy defect with removal of the orbital floor and periorbita. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
hypocapnia through controlled hyperventilation, mannitol diuresis, or steroids. This also lessens the need for brain retraction, which minimizes postoperative brain edema. Next, the dura is carefully dissected along the floor of the anterior cranial fossa to expose the crista galli and olfactory grooves. The olfactory nerves are transected to expose the cribriform plate. Dural elevation is continued to expose the fovea ethmoidalis and orbital roofs. Posteriorly, the planum sphenoidale and the base of the anterior clinoid process may be exposed as dictated by the extent of the tumor. If the dura is
involved by the tumor, intradural exposure is achieved and dura incisions are made around the tumor and then the dissection proceeds in a subdural plane and the dura and even brain tissue, if involved, is resected along with the tumor. With simultaneous exposure provided superiorly through the intracranial approach and inferiorly through the extracranial approach, osteotomies of the cranial floor around the tumor can be safely completed. Malleable retractors are used to protect the brain and the orbit as osteotomies are made. The placement of osteotomies and the extent of
254
Hanna et al.
Foramen rotundum Periorbita
V2 Pterygopalatine ganglio; fossa Bur Maxillary sinus
Zygomatic bone
(A)
Periorbital membrane
Sphenoid sinus Suction-showing dissection of temporal dura from V2 and the trigeminal ganglion
Internal acrotid a.
Temporal dura
Nasal cavity
(B)
(C)
Figure 30 Dissection of the foramen rotundum through the transmaxillary route (A) and close-up schematic (B). Intraoperative view (C) of thickened V2 within the foramen rotundum (arrow). Source: This illustration in the property of The Department of Neurosurgery, M.D. Anderson Cancer Center is used with permission.
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
255
Figure 31 Anterior craniofacial resection: Extracranial approaches. In addition to the frontal craniotomy, the extracranial approach may be transfacial (A), sublabial (B), or endonasal (C). Source: This illustration in the property of The Department of Neurosurgery, M.D. Anderson Cancer Center is used with permission.
resection are dictated by the extent of tumor involvement, and tailored in each case. Typically, however, osteotomies are made from the planum sphenoidale, along the roof of the ethmoid, and forward to the front of the cribriform plate (Fig. 36). Frozen section control of the margins should be done to ensure the adequacy of resection.
Figure 33 Subfrontal approach. Note the increased basal exposure provided by the subfrontal approach by incorporating the supraorbital rim and glabella and nasal bones with the frontal craniotomy. Source: This illustration in the property of The Department of Neurosurgery, M.D. Anderson Cancer Center is used with permission.
(A)
(B)
(C)
(D)
(E)
(F)
Figure 32 Frontal Craniotomy. (A, B) Bicoronal incision. (C) Incision of the superficial layer of the deep temporalis fascia. Further dissection is done deep to this plane to preserve the frontal branch of the facial nerve. (D) The pericranial flap. Adequate length and good vascular supply of the flap are prerequisites for effective reconstruction of the skull base defect. (E) Frontal craniotomy (schematic). (F) Frontal craniotomy (intraoperative photograph). Source: This illustration in the property of The Departments of Head and Neck Surgery (A–D, F) Neurosurgery (E), M.D. Anderson Cancer Center is used with permission.
256
Hanna et al.
(A)
resection (12). Even with the best imaging techniques, the definitive and most accurate assessment of the extent of orbital invasion, and whether the eye could be preserved, has to be made intraoperatively. This needs to be clearly discussed with the patient and family, and an informed consent for possible exenteration needs to be obtained in high-risk cases. There is an evolving role for induction chemotherapy and concurrent chemoradiation in the management of patients with orbital invasion by advanced sinonasal cancers (Fig. 37). The role of such neoadjuvant treatment in enhancing the chances of orbital preservation continues to be investigated (19). If a decision is made to exenterate the orbit, supraand subciliary incisions are made around the upper and lower eyelids, respectively. This allows for preservation of the
(B)
Figure 34 Pre- (A) and postoperative (B) MRI of tumor involving the planum sphenoidale and upper two-thirds of the clivus removed via a subfrontal approach. Source: This illustration in the property of The Department of Neurosurgery, M.D. Anderson Cancer Center is used with permission.
Management of the Orbit Sinonasal Tumors Every effort should be made to preserve the eye as long as preservation does not compromise the adequacy of oncologic resection. Attempts at orbital preservation in the face of gross residual disease, however, usually result in poor disease control and ultimate loss of orbital function. Most studies have shown that if orbital invasion is limited to the bony orbit or the periorbita, orbital preservation is possible without compromising oncologic outcome (1). Orbital exenteration is usually indicated when there is gross invasion of the periorbital fat, extraocular muscles, or optic nerve. The presence of proptosis or diplopia may be due to displacement rather than invasion of the intraorbital contents. Decreased visual acuity or visual fields, or the presence of an afferent pupillary defect, usually indicate gross invasion of the orbit. Orbital invasion by perineural spread rather than direct extension may be missed unless careful examination of the cranial nerves, especially V1 and V2, is done. Detailed neuro-ophthalmologic examination should be conducted on all patients with suspected or confirmed orbital involvement by sinonasal or other skull base tumors. If orbital exenteration is contemplated, always make sure that the patient has useful vision in the contralateral eye. In the absence of any ocular signs or symptoms, however, evaluation of the extent of orbital involvement relies mainly on imaging. High-resolution CT and MRI are complimentary and provide critical information regarding the extent of orbital bony and soft tissue involvement, respectively. CT scans obtained at 1- to 3-mm slices with detailed bone windows are best for evaluating bony involvement of the orbital walls. MRI is best used to evaluate the extent of soft tissue invasion beyond the periorbita (Fig. 37). MRI is also useful in detecting perineural spread proximally beyond the orbital apex and into the cavernous sinus or optic chiasm, which compromises surgical margins, local disease control, and survival, and as such is a contraindication for surgical
(D)
(E)
Figure 35 Subfrontal approach. (A) Bicoronal incision and soft tissue dissection. The scalp flap is reflected anteriorly down to the level of the nasal bones. The supraorbital nerves were surrounded by complete foramina rather than a notch. Osteotomies around the foramina allow downward reflection of the nerves with the soft tissue flap. Burr holes are placed on either side of the SSS, anterior to the coronal suture. (B) Bilateral burr holes are placed posterior to the frontal-zygomatic sutures. These anatomic keyholes provide access to the anterior fossa dura and periorbita, separated by the bony orbital roof. (C) Orbital and nasal osteotomies. (D, E) The cranio-orbital– nasal bone flap is removed as a single unit as illustrated by the diagram (D) and intraoperative photograph (E). Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
257
(A)
(B)
(C)
Figure 36 Resection of the floor of the anterior cranial fossa. (A, B) Osteotomies along the floor of the anterior cranial fossa. (C) Intraoperative photograph showing resection of the floor of the anterior cranial fossa and the medial orbital walls. Source: This illustration in the property of The Department of Neurosurgery (A andB) and Head and Neck Surgery (C), M.D. Anderson Cancer Center is used with permission.
eyelids, which can be used to line the orbit. If the eyelids are involved with cancer, they must be included in the resection (Fig. 38). The periorbita is incised over the superior and lateral orbital rims. Dissection continues along the roof of the orbit and lateral walls, until the SOF and the optic foramen are exposed. Lidocaine is injected around these structures to block any autonomic-induced cardiac arrhythmias. To prevent troublesome bleeding, the neurovascular structures in the SOF are slowly and carefully isolated, ligated or clipped, and transected. The optic nerve and the ophthalmic artery are then managed in a similar fashion. The extraocular muscles are transected at their origin in the orbital apex. The medial
and inferior orbit may be left attached to the specimen if en bloc resection of the eye in patients with sinonasal cancer is indicated. Osteotomies are done, as previously described, for total maxillectomy, except that the orbital bony cuts are connected at the orbital apex, rather than the inferior orbital fissure. Primary Orbital Tumors Surgery for primary orbital tumors will depend on whether the tumor is benign or malignant. Examples of benign tumors include vascular malformation, neurofibroma, osteoma, and pleomorphic adenoma of the lacrimal gland
258
Hanna et al.
(Figs. 39–42). Examples of malignant tumors include lymphoma, which is the most common, and salivary gland tumors of the lacrimal gland such as adenoid cystic carcinoma (Fig. 43). In most cases of primary intraorbital tumors, the surgical approach is largely dictated by the location of the tumor. Tumors of the medial orbit may be removed through a transethmoidal modified Lynch approach (Fig. 41). The transconjunctival approach, with detachment of the lateral canthal tendon, provides access to the orbital floor. Tumors of the lateral orbit are best approached through a lateral orbitotomy, which can be done using either a brow incision (modified Kronlein or Stallard-Wright incision) or a hemi- or bicoronal approach. The advantages of the hemi- or bicoronal approaches include wider exposure for tumor resection, avoiding periocular incisions and excellent cosmetic result (Fig. 42). Tumors of the superior orbit or tumors invading intracranially are best resected through a cranio-orbito– zygomatic approach (Fig. 43).
(A)
Reconstruction, Complication Avoidance, and Rehabilitation
(B)
Figure 37 Sinonasal undifferentiated carcinoma with orbital invasion. (A) MRI showing tumor of the right sinonasal region with gross invasion of the medial and inferior orbital contents. The tumor also invades the anterior skull base with limited intracranial extension. (B) MRI of same patient 3 years after completing treatment with induction chemotherapy followed by concurrent chemoradiation. Imaging shows no evidence of residual disease. Surgery was avoided and the orbit was preserved. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
(A)
(B)
Meticulous reconstruction of surgical defects resulting from resection of sinonasal and orbital tumors is essential to preventing complications and optimizing surgical and functional outcome. The goals and strategies of reconstruction will depend on the extent and location of the surgical defect. Table 3 summarizes the major areas of reconstruction and their respective goals. Palatal Reconstruction: Oronasal Separation Effective separation between the oral and nasal cavities is essential for effective speech and deglutition. Palatal defects resulting from a maxillectomy are simply and effectively sealed using a prosthetic obturator. Preoperatively, the oromaxillofacial prosthodontist takes dental impressions,
(C)
Figure 38 Orbital exenteration. (A) Circumorbital incisions. (B) Tumor specimen, including en bloc resection of the maxilla and orbit. (C) Immediate postoperative appearance. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
(A)
259
(B)
Figure 39 Orbital neurofibroma of the abducent nerve. (A) Axial CT scan and T1-MRI with gadolinium (B) showing the tumor following the course of the abducent nerve in the lateral orbit and extending to the cavernous sinus. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
and designs a surgical obturator, which is used at the end of surgery to seal the palatal defect (Fig. 44) (20). The surgical obturator can be slightly modified intraoperatively to custom-fit the defect. The advantages of this immediate reconstruction are early postoperative restoration of normal speech and oral feeding. This minimizes the early postoperative morbidity of surgery, and obviates the need for enteral feeding. Preoperatively, a clear communication between the head and neck surgeon and the maxillofacial prosthodontist concerning the anticipated maxillary defect is required for optimal results. An additional advantage of the surgical obturator is its ability to support the surgical packing used to immobilize the skin graft lining the cheek flap. This epithelial lining, when completely healed, minimizes granulation tissue formation, provides a smooth mucosal lining, reduces scar contracture of the cheek, provides a scar band to support the obturator, and facilitates
(A)
cavity hygiene [Fig. 44(A)]. At the end of the first postoperative week, and after removal of the surgical packing, the surgical obturator is replaced with an interim obturator [Fig. 44(B)]. This is used for several weeks after discharging the patient from the hospital and after completion of adjuvant therapy, allowing the surgical cavity to heal completely. Patients are instructed to remove the obturator periodically and clean the cavity with saline irrigation. At follow-up visits, the obturator is removed, and cleaned of any crusts or debris. Finally, a permanent obturator is designed to custom-fit the cavity after it matures to its final shape and dimensions. The acrylic dome incorporated into its design provides some cheek support [Fig. 44(C) and 44(D)]. In addition to being a simple and effective method of reconstruction of palatal defects, permanent obturators can also provide full dental restoration. At follow-up visits, removal of the obturator allows for
(B)
Figure 40 Pleomorphic adenoma of the lacrimal gland. This young woman presented with proptosis and a palpable mass in the upper lateral right orbit (A). A coronal MRI (B) shows the mass in the region of the right lacrimal gland. Surgical resection via a lateral orbitotomy revealed a pleomorphic adenoma. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
260
Hanna et al.
(A)
(E)
(B)
(F)
(G)
(C)
(D)
(H)
Figure 41 Osteoma of the ethmoid and medial orbital wall. (A) Coronal CT showing an osteoma of the left ethmoid and medial orbital wall. (B) “Lynch” incision for an anterior ethmoidectomy. (C) Exposure of the tumor in the medial orbit. (D) Bony dissection around the tumor. (E) Surgical specimen. (F) Surgical field showing complete resection. (G) Closure. (H) Postoperative appearance. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
easy inspection of the cavity for any evidence of recurrent disease. Surgical reconstruction of palatal defects with tissue flaps has the advantage of eliminating the need for regular hygiene of the surgical cavity. Flap reconstruction, however, requires more extensive surgery, adds donor site morbidity, does not allow rapid dental restoration, and conceals the surgical cavity from inspection for recurrent tumor. These disadvantages make prosthetic obturators the method of choice for reconstruction of surgical defects of the palate. The use of tissue flaps is usually reserved to patients who need additional reconstruction to the maxillary skeleton, orbital floor, or cranial base. In these cases, osseointegrated implants are used to facilitate dental restoration (see Fig. 2 in chap. 9, “Prosthetic Rehabilitation of Patients undergoing Skull Base Surgery”). Cranial Base Reconstruction: Cranionasal Separation Whenever the cranial and nasal cavities are joined by a surgical defect, as is the case with anterior craniofacial resection, watertight cranionasal separation is essential to reduce the risk of CSF leak, meningitis, and pneumocephalus. Meticulous closure of all dural incisions is necessary. Larger defects of the dura should be repaired using temporalis fascia, pericranium, or fascia lata grafts. Although bony reconstruction of the anterior skull base has been described using vascularized and nonvascularized bone grafts as well as bone cement,
reconstruction of the bone defect is not routinely necessary in most patients. The vascularized pericranial flap is currently the most frequently used flap for reconstructing defects of the floor of the anterior cranial fossa [Fig. 32(D)]. Flap handling and suturing should be meticulous in order to achieve a watertight seal (Fig. 45). Fibrin glue and tissue adhesives do not compensate for an imperfect closure. Lumbar subarachnoid drainage may be used for several days postoperatively to reduce CSF pressure and the possibility of a leak. Excessive lumbar drainage, however, may encourage the development of pneumocephalus. Occasionally, more bulk is needed to reconstruct the surgical cavity and reduce dead space, such as with extensive defects of the cranial base. Regional flaps, such as the temporalis muscle, are usually adequate for this purpose. If the muscle bulk is inadequate, or if its blood supply has been sacrificed, then a microvascular free flap is used (21). Table 3
Reconstruction of Surgical Defects
Defect
Goals of reconstruction
Palate Cranial base Orbital-maxillary Dental Facial
Oronasal separation Cranionasal separation Eye and cheek support Restoration of dentition and occlusion Restoration of facial defects and cosmesis
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
(A)
(B)
(E)
(C)
(F)
(H)
261
(D)
(G)
(I)
Figure 42 Bicoronal approach for lateral orbitotomy. (A) Preoperative axial T2-MRI showing a vascular malformation of the left orbit. (B) Postoperative axial T2-MRI showing complete resection of the tumor. (C–G) Intraoperative photographs showing bicoronal approach (C), orbito-zygomatic exposure and osteotomies (arrows) (D), orbito-zygmoatic bone segment including the lateral orbital rim and zygomatic arch (E), lateral orbitotomy provides wide exposure for tumor resection (F), and tumor specimen (G). (H, I) Postoperative photographs showing good eye position and cosmetic result. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
Vascularized tissue may also be needed to protect the carotid artery if it is exposed to the surgical defect. This is done to prevent desiccation of the arterial wall and carotid artery blowout. This is particularly important if the patient received prior radiation therapy or will receive postoperative adjuvant radiation (see chap. 8, “Surgical Reconstruction of Skull
Base Defects,” for more detail on microvascular reconstruction lease). Orbito-Maxillary Reconstruction: Eye and Cheek Support The maxilla has three bony buttresses: the nasomaxillary, zygomatico-maxillary, and pterygo-maxillary. In addition to
262
Hanna et al.
(A)
(C)
(B)
(D)
Figure 43 Adenoid cystic carcinoma of the lacrimal gland. Coronal CT (A) and MRI (B) showing bone destruction of the superior orbital wall and intracranial soft invasion, respectively. The tumor was completely resected through a cranio-orbito–zygomatic approach (C), and the superolateral orbital walls were reconstructed using porous polyethylene “Medpore” implant (D). Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
a palatal defect, total maxillectomy results in resection of all three buttresses and loss of adequate skeletal support to the soft tissues of the cheek [Fig. 29(D)]. Loss of the zygomaticomaxillary buttress results in inferior displacement of the orbit and flattening of the malar eminence. Loss of the nasomaxillary buttress results in superior and posterior deviation of the alar base of the nose. Loss of the pterygomaxillary buttress results in superior and posterior deviation of the upper lip. Resection of the inferior orbital rim and floor leads to loss of skeletal support to the eyelid and globe. The lack of adequate orbital support leads to enophthalmos, ectropion, hypoglobus, and diplopia resulting in unacceptable esthetic and functional outcome. The combined effects of loss of support of the eye and cheek result in the “typical” postmaxillary deformity [Fig. 46(A)]. Although a split thickness
skin graft and a palatal obturator are commonly used in reconstruction after maxillectomy, these methods of reconstruction do not provide adequate support for the cheek and eye. Skeletal reconstruction of the maxillary and orbital defects may be done using bone grafts, most commonly autogenous calvarial grafts. Demineralized and banked bone grafts, which are available in a variety of shapes and sizes, may also be used. Alternatively, alloplastic implants may be used for reconstruction of the maxillary buttresses, orbital rim, or orbital floor. Titanium mesh, bone cement, and porous polyethylene have all been successfully used in orbital and maxillary bony reconstruction. Whatever method is used for bone reconstruction, adequate coverage with well-vascularized soft tissue is essential
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
(A)
(C)
263
(B)
(D)
Figure 44 Prosthetic rehabilitation of the palate. (A) Palatal defect resulting from maxillectomy with split thickness skin graft lining the surgical cavity. (B) Interim obturator effectively seals the palatal defect and allows patients to resume a soft diet. (C, D) Permanent obturator provides excellent palatal reconstruction and dental restoration, allowing the patient to resume a normal diet. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
(A)
(B)
Figure 45 Sagittal (A) and (B) axial views of pericranial flap reconstruction of anterior cranial base defect. Source: This illustration in the property of The Department of Neurosurgery, M.D. Anderson Cancer Center is used with permission.
264
Hanna et al.
(A)
(B)
Figure 46 Orbito-maxillary reconstruction. (A) Left total maxillectomy with no reconstruction. “Typical” postmaxillectomy deformity due to lack of support of the orbit and cheek. There is downward displacement of the orbit and medial canthus. Note the presence of enophthalmos as evidenced by a prominent upper eyelid sulcus. There is flattening of the cheek and deviation of the nasal tip. The patient had significant diplopia. These deformities can be avoided with adequate bony reconstruction. (B) Left total maxillectomy with orbito-maxillary reconstruction. Postoperative appearance showing good position of the eye, cheek, and nose. Source: This illustration in the property of The Department of Head and Neck Surgery, M.D. Anderson Cancer Center is used with permission.
to prevent resorption of bone grafts and infection or extrusion of alloplastic implants. The pedicled temporalis muscle flap, temporoparietal fascial flap, or septal mucosal flap are most commonly used for this purpose. Microvascular free flaps may be used to either provide soft tissue coverage of bone grafts or implants, or composite vascularized bone flaps may be used for full reconstruction of both soft tissue and bone defects (21). Primary reconstruction of defects of the maxilla and orbit at the time of maxillectomy is easier and results in better esthetic and functional outcome than delayed reconstruction [Fig. 46(B)]. Although secondary reconstruction after globesparing maxillectomy is feasible, it is often difficult and the results are limited by excessive scarring and soft tissue contracture, especially in patients who underwent adjuvant radiation therapy. These patients may benefit more from freetissue transfer reconstruction (21).
Dental Restoration Prosthetic rehabilitation using partial or full upper dentures is the easiest method of dental restoration in patients who have undergone maxillectomy (Fig. 44). Remaining contralateral teeth facilitate retention of partial dentures. In edentulous patients, dental fixatives can be used, but denture retention is more difficult. A soft palate “band” at the posterior edge of the defect may provide enough of a ledge for retaining the prosthesis. Retention of the prosthesis in edentulous patients who underwent an extended resection to include the soft palate may be extremely difficult. In such cases, the use of osteointegrated implants facilitates prosthetic dental restoration (see Fig. 2 in chap. 9, “Prosthetic Rehabilitation of Patients undergoing Skull Base Surgery”).
Another important aspect in the rehabilitation of patients after maxillectomy is the prevention of trismus. Patients who have undergone resection of the pterygoid plates or muscles of mastication are particularly prone to develop trismus. This may be severe enough to interfere with inserting and wearing a denture-bearing obturator. Early postoperative jaw-opening exercises using “stacked” wooden tongue blades or commercially available devices (e.g., Therabite) are extremely important in preventing or minimizing postoperative trismus. Restoration of Facial Defects Smaller defects of the face are optimally reconstructed using local flaps. Local skin flaps provide the best thickness and color match for facial defects. Reconstruction of more extensive facial defects may require regional or microvascular free flaps, but suffer from less optimal esthetic outcome due to color or thickness mismatch between the donor and defect sites. Facial defects resulting from orbital exenteration or total rhinectomy are best managed with prosthetic restoration (see Figs. 3 and 4 in chap. 9, “Prosthetic Rehabilitation of Patients undergoing Skull Base Surgery”).
SUMMARY Surgical management of tumors of the nasal cavity, paranasal sinuses, and orbit require in depth understanding of the biologic behavior of the various tumor types originating in this region. Careful assessment of the location and extent of the tumor is crucial in planning and executing the proper surgical approach for adequate oncologic resection. Meticulous reconstruction of the surgical defect(s) is critical in avoiding
Chapter 16: Surgical Management of Tumors of the Nasal Cavity, Paranasal Sinuses, Orbit, and Anterior Skull Base
complications and maximizing functional and esthetic outcome. As with other types of cranial base tumors, multidisciplinary collaboration is essential for successful treatment of patients with tumors of this region.
REFERENCES 1. Hanna EY, Westfall CT, Myers EN, et al. Cancer of the nasal cavity, paranasal sinuses, and orbit. In: Myers EN, Suen JY, Myers JN, Hanna EY, eds. Cancer of the Head and Neck. Philadelphia, PA: Saunders, 2003:155–206. 2. Diaz EM Jr, Johnigan RH III, Pero C, et al. Olfactory neuroblastoma: The 22-year experience at one comprehensive cancer center. Head Neck. 2005;27(2):138–149. 3. Rosenthal DI, Barker JL Jr, El-Naggar AK, et al. Sinonasal malignancies with neuroendocrine differentiation: Patterns of failure according to histologic phenotype. Cancer. 2004;101(11):2567– 2573. 4. Girod D, Hanna E, Marentette L. Esthesioneuroblastoma. Head Neck. 2001;23(6):500–505. 5. Lupinetti AD, Roberts DB, Williams MD, et al. Sinonasal adenoid cystic carcinoma: The M. D. Anderson Cancer Center experience. Cancer. 2007;110(12):2726–2731. 6. Prokopakis EP, Snyderman CH, Hanna EY, et al. Risk factors for local recurrence of adenoid cystic carcinoma: The role of postoperative radiation therapy. Am J Otolaryngol. 1999;20(5):281–286. 7. Pitman KT, Prokopakis EP, Aydogan B, et al. The role of skull base surgery for the treatment of adenoid cystic carcinoma of the sinonasal tract. Head Neck. 1999;21(5):402–407. 8. Tirado Y, Williams MD, Hanna EY, et al. CRTC1/MAML2 fusion transcript in high grade mucoepidermoid carcinomas of salivary and thyroid glands and Warthin’s tumors: Implications for histogenesis and biologic behavior. Genes Chromosomes Cancer. 2007;46(7):708–715. 9. Hanna E, Vural E, Teo C, et al. Sinonasal tumors: The Arkansas experience. Skull Base Surg. 1998;8(Suppl):15.
265
10. Hanna E, Linskey ME, Pieper D. Malignant tumors of the anterior cranial base. In: Sekhar LN, Fessler RG, eds. Atlas of Neurosurgical Techniques. New York: Thieme, 2006:588–598. 11. DeMonte F, Ginsberg LE, Clayman GL. Primary malignant tumors of the sphenoidal sinus. Neurosurgery. 2000;46(5):1084– 1091; discussion 1091–1092. 12. Hanna E, Vural E, Prokopakis E, et al. The sensitivity and specificity of high-resolution imaging in evaluating perineural spread of adenoid cystic carcinoma to the skull base. Arch Otolaryngol Head Neck Surg. 2007;133(6):541–545. 13. Cohen ZR, Marmor E, Fuller GN, et al. Misdiagnosis of olfactory neuroblastoma. Neurosurg Focus. 2002;12(5):e3. 14. Bristol IJ, Ahamad A, Garden AS, et al. Postoperative radiotherapy for maxillary sinus cancer: Long-term outcomes and toxicities of treatment. Int J Radiat Oncol Biol Phys. 2007;68(3):719–730. 15. Hanna EY, Holsinger C, Demonte F, et al. Robotic endoscopic surgery of the skull base: A novel surgical approach. Arch Otolaryngol Head Neck Surg. 2007;133(12):1209–1214. 16. Vural E, Hanna E. Extended lateral rhinotomy incision for total maxillectomy. Otolaryngol Head Neck Surg. 2000;123(4):512–513. 17. DeMonte F, Hanna E. Transmaxillary exploration of the intracranial portion of the maxillary nerve in malignant perineural disease. Technical note. J Neurosurg. 2007;107(3):672–677. 18. Feiz-Erfan I, Suki D, Hanna E, et al. Prognostic significance of transdural invasion of cranial base malignancies in patients undergoing craniofacial resection. Neurosurgery. 2007;61(6):1178– 1185; discussion 1185. 19. Essig GF, Newman SA, Levine PA. Sparing the eye in craniofacial surgery for superior nasal vault malignant neoplasms: Analysis of benefit. Arch Facial Plast Surg. 2007;9(6):406–411. 20. Dexter WS, Jacob RF. Prosthetic rehabilitation after maxillectomy and temporalis flap reconstruction: A clinical report. J Prosthet Dent. 2000;83(3):283–286. 21. Chang DW, Langstein HN, Gupta A, et al. Reconstructive management of cranial base defects after tumor ablation. Plast Reconstr Surg. 2001;107(6):1346–1355; discussion 1356–1357. 22. Gray H. Anatomy of the Human Body, 20th ed., edited by Lewis WH, New York: Bartleby, 2000.
17 Tumors of the Nasopharynx William Ignace Wei and Paul K. Y. Lam
The lymphatic supply of the nasopharynx is found mainly in the submucosal region, which drains into the retropharyngeal lymph nodes. Efferents lymphatics from these nodes and some of the lymphatics that come directly from the nasopharynx drain to the deep cervical lymph nodes. The lymphatic drainage through the neck nodes goes in an orderly fashion, from upper to lower cervical lymph nodes.
SURGICAL ANATOMY The pharynx is divided into three parts, the nasopharynx, oropharynx, and hypopharynx. The nasopharynx is the upper portion of the pharynx and is separated from the oropharynx by the soft palate. Anatomically, it is the space located behind the nasal cavities and its mucosal lining starts at the posterior choana. The undersurface of the body of the sphenoid bone forms the slanting roof of the nasopharynx that continues inferiorly into the posterior wall, which is formed by the arch of the atlas vertebra and upper part of the body of the axis vertebra. The floor of the nasopharynx is formed by the upper surface of the soft palate, which also separates the nasopharynx from the oropharynx below. The lateral wall of the nasopharynx is formed by the opening of the Eustachian tube superiorly and the upper part of the superior constrictor muscle inferiorly. The orifice of the Eustachian tube is guided by an incomplete cartilaginous ring, which is deficient in the inferolateral aspect. The medial portion of the incomplete cartilaginous ring elevates the overlying mucosa to form the medial crura. The slit-like space situated medial to this crura ¨ is the fossa of Rosenmuller; its size and depth varies between individuals. The superior constrictor forms the muscular layer of the nasopharynx starting from beneath the auditory tympanic tube, and investing this muscle externally is the pharyngobasilar fascia. This fascia joins its counterpart from the opposite side to form the median raphe, which extends from the skull base caudally throughout the posterior pharyngeal wall. The pharyngobasilar fascia together with the prevertebral fascia encloses the retropharyngeal space, which harbors the lymph node of Rouviere. Lateral to the superior constrictor is the parapharyngeal space, which contains the last four cranial nerves, the carotid sheath, and the sympathetic trunk. The superior wall of the nasopharynx is lined by pseudostratified ciliated epithelium near the choana. The posterior wall is lined with stratified squamous cells and the epithelium lies on a well-defined basement membrane. On the lamina propria, abundant lymphoid tissue can be found. This in the midline forms the pharyngeal tonsil or the adenoid tonsil, which is prominent in children. This lymphoid aggregate is a part of the Waldeyer ring of lymphatic tissue guarding the pharynx. The arterial blood supply of the nasopharynx comes from branches of the internal maxillary artery, while the venous drainage of the region goes to the pterygoid plexus and then onto the facial and the internal jugular veins. The nerve supply of the nasopharynx comes from the pharyngeal plexus, although the nasopharynx is immobile. The sensory nerve supply of the region comes from the branches of the maxillary nerve.
REGIONAL PATHOLOGY AND DIFFERENTIAL DIAGNOSIS Tumors in the nasopharynx include congenital, benign, and malignant neoplasms. Congenital tumors and tumor-like conditions are in general uncommon and include dermoids, teratomas, and encephalocele. These pathologies are usually seen in neonates or young children and they present with nasal obstruction or breathing problems. Subsequent investigations such as imaging studies or endoscopic examinations will give clues to the diagnosis.
Benign Tumors Benign tumors arising from the epithelial linings of the nasopharynx are rare. Other benign tumors include benign salivary gland tumors arising from the minor salivary glands and juvenile angiofibroma. Juvenile angiofibroma is a benign but locally aggressive tumor affecting most commonly adolescent boys. Symptoms include nasal obstruction and epistaxis, which may be severe. This tumor arises from the posterolateral aspect of the roof of the nasal cavity in the region of the sphenopalatine foramen (1). The tumor grows into the pterygopalatine fossa and then posteromedially into the nasopharynx or laterally to the infratemporal fossa. It may also extend superiorly eroding into the sphenoid sinus or anteriorly to affect the maxillary sinus. It may extend into the orbit through the inferior orbital fissure leading to proptosis and further invasion superiorly will go to the middle cranial fossa as well. Macroscopically, the tumor is lobulated with the consistency ranging from spongy to firm depends on the proportion of vascular tissue and fibrous component. Histologically, the tumor is noncapsulated and consists of numerous blood vessels of varying caliber coursing through a fibrous tissue stroma. The thickness of the muscular coat of these vessels varies and there is a lack of elastic fibers contributing to the frequent episodes of epistaxis. These clinical features frequently lead to the correct diagnosis of this condition, while imaging studies help to confirm it. Plain radiographs demonstrate the bowing of the anterior wall of the maxillary sinus anteriorly and the erosion of the base of the medial pterygoid plate. Angiography reveals a typical vascular tumor blush with the internal maxillary artery as the principal feeder. 267
268
Wei and Lam
Malignant Tumors Adenoid cystic carcinoma and adenocarcinoma arising from the minor salivary glands situated in the nasopharynx are uncommon. Grossly, they may present as a polypoid mass (Fig. 1) or as a submucosal bulge in the nasopharynx. Symptoms associated with these tumors are usually trivial until they infiltrate and affect the surrounding structures such as the fifth cranial nerve. Then the patient may present with facial pain and trismus. Biopsy of the tumor confirms the diagnosis. Sarcoma of the nasopharynx is rarely seen and includes rhabdomyosarcoma in children and malignant fibrous histiocytoma in adults. The frequently encountered symptoms are nasal obstruction, epistaxis, and serous otitis media. The gross appearance is the presence of a firm smooth mass filling the nasopharynx and the exact diagnosis is confirmed by biopsy. Chordomas are malignant neoplasms arising from the remnant of notochord tissue located at the spheno-occipital region. The characteristic location of a cranial chordoma is at the clivus as it forms the posterior wall of the nasopharynx; many patients with chordoma present with a mass in the nasopharynx (Fig. 2). The symptoms are related to the cranial nerves involved and the effect of the tumor bulk. Extramedullary plasmacytoma are neoplasms that arise from the plasma cells located in the extraosseous soft tissue. This is more frequently seen in middle-aged or elderly men. The symptoms are related to the presence of the soft tissue tumor in the nasopharynx which might fill the entire nasopharynx and there is a tendency of metastasis to regional lymph nodes. As the lymphoid tissue in the nasopharynx is part of the Waldeyer ring, one of the differential diagnoses of malignant lesion in the region is malignant lymphoma. The majority of these are diffuse B cell lymphomas and the symptoms are again related to the presence of the mass in the nasopharynx causing nasal obstruction or impairing function of the Eustachian tube. Other malignant pathologies found in the nasopharynx are lesions in the nasal cavity that extend to the nasopharynx. These include the sinonasal carcinomas and malignant mucosal melanomas arising from the nasal lining. The common symptoms are nasal obstruction and epistaxis related to the tumor bulk and the superficial ulceration.
Figure 2 Plain lateral X-ray showing a chordoma producing a bulge at the posterior wall of the nasopharynx (arrow).
For inhabitants of southern China, northern Africa, and Alaska, the commonest malignancy in the nasopharynx is nasopharyngeal carcinoma (NPC). This carcinoma arises from epithelial lining of the nasopharynx, most frequently from ¨ the fossa of Rosenmuller, the recess located medial to the medial crura of the Eustachian tube. NPC is a squamous cell carcinoma with varying degree of differentiation and among the majority of the Chinese patients—these are undifferentiated squamous cell carcinoma (2). These malignant squamous cells are frequently intermingled with lymphoid cells in the nasopharynx, thus it was previously termed as lymphoepithelioma. Electron microscopy studies have, however, confirmed the squamous origin of these malignant cells. Patients suffering from NPC may present with one or more of the four groups of symptoms. Epistaxis, nasal obstruction, and discharge are related to the presence of tumor mass in the nasopharynx. When the tumor affects the Eustachian tube, the patient presents with hearing loss and tinnitus. Serous otitis media was noted in 40% of the newly diagnosed NPC patients. Thus when a Chinese adult patient presents with serous otitis media, then the possibility of NPC should be excluded (3). With the superior extension of the tumor to erode the skull base, the patient may present with headache and the tumor may involve the third, fourth, fifth, and sixth cranial nerves. Commonly, patients may present with enlarged cervical lymph nodes, which usually appear at the upper neck (Fig. 3). As the nasopharynx is located in the midline, it may metastasize to the opposite side, thus it is not uncommon to see patients presenting with bilateral enlarged cervical lymph nodes.
CLINICAL ASSESSMENT
Figure 1 Flexible endoscopic view of an adenocarcinoma in the nasopharynx (arrow). Left Eustachian tube cartilage is also shown (C).
For patient suspected of harboring pathology in the nasopharynx, a full clinical examination should be carried out to evaluate the general condition of the patient and to identify any evidence of distant metastasis. The age and medical history of the patient might give clues to the diagnosis. Comprehensive clinical evaluation of the head and neck region is also essential. The test of nasal obstruction either one side or bilaterally should be carried out and an
Chapter 17: Tumors of the Nasopharynx
269
Figure 4 A 0-degree rigid endoscopic view of an NPC (T) in the nasopharynx, located medial to the medial crura (C) of right Eustachian tube opening.
Figure 3 Clinical photograph of a patient suffering from NPC, presenting with cervical lymph node.
indirect examination of the nasopharynx with a mirror may reveal important information toward the diagnosis. The anterior nasal cavity should be examined for macroscopic tumor and lesions arising from the mucosa. The ears should be examined for hearing impairment and the presence of serous otitis media. The neck should always be palpated for the presence of metastatic cervical lymph node.
can be inserted through this channel to take a biopsy of the tumor under direct vision. It allows thorough examination of the entire nasopharynx when it is inserted through the nasal cavity. It may also be manipulated upward behind the soft palate to examine the nasopharynx from the inferior aspect. In view of the size of the endoscope and its flexibility, it is well tolerated by most patients. Despite all these advantages, the visual image gathered with the flexible endoscope is inferior to that of the rigid telescope. The cup size of the biopsy forceps is limited by the size of the channel on the endoscope, thus the amount of tissue obtained for histological examination might not be adequate. When biopsy is taken with the flexible endoscope, then multiple biopsies might have to be taken to increase the yield. Sometimes, NPC may be submucosally located in the submucosa, thus the mucosal surface has to be broken with the forceps to enable deeper tissue to be obtained (4).
Endoscopic Examination The nasal cavity and the nasopharynx can be examined with either the fiber-optic endoscope or the rigid telescopes under topical anesthesia. The rigid Hopkins telescopes are 2.7 or 4 mm in diameter with different angles of viewing. Both the 0-degree and the 30-degree telescopes give an excellent view of the nasopharynx (Fig. 4) and the location of the tumor in the nasopharynx. A 70-degree endoscope inserted behind the uvula allows visualization of the roof of nasopharynx and both Eustachian tube openings (Fig. 5). It gives an idea of the extension of the tumor across the midline and also the vascular nature of the tumor. These rigid telescopes, however, do not have a channel for suction or for the introduction of biopsy forceps. When a biopsy is needed, the biopsy forceps has to be inserted along the side of the endoscope. A relatively large biopsy forceps can be inserted to obtain an adequate amount of tumor for histological examination. The fiber-optic flexible endoscope for ordinary use has a diameter of 6 mm, it has a suction channel for the removal of nasal secretions during examination, and a biopsy forceps
Figure 5 A 70-degree endoscope inserted behind the uvula and the NPC (T) can be seen together with posterior edge of nasal septum (N) and the opposite Eustachian tube opening (arrow).
270
Wei and Lam
and its vicinity. The neck should also be included in the imaging studies for the detection of occult cervical metastasis. These cross-sectional imaging studies provide information on the deep extension of the tumor including skull base erosion and extension of the lesion intracranially.
Computed Tomography Besides demonstrating the soft tissue extent of the tumor in the nasopharynx and its extension laterally into the paranasopharyngeal space (8), CT is also sensitive in detecting bone erosion at the skull base (Fig. 6). Tumor extension intracranially through the foramen ovale via perineural spread along the third division of the trigeminal nerve (V3) and even to the cavernous sinus should be evaluated (9). CT in the posttherapy phase may show bone regeneration and this indicates complete eradication of tumor (10).
Magnetic Resonance Imaging Figure 6 Direct coronal CT showing tumor in the paranasopharyngeal space eroding the skull base (arrow).
Serological Tests Serological tests are not applicable for most malignant lesions situated in the nasopharynx. It is applicable when the symptoms suggest that the patient may be harboring NPC, then their antibody titer, the immunoglobulin A, in response to the early antigen and the viral capsid antigen of Epstein Barr virus may further indicate the presence of the disease (5). Recently, cell-free DNA of the Epstein Barr virus (EBV) has been evaluated as a tumor marker for detection of NPC (6), yet its application in the detection of small recurrent tumor after radiotherapy is limited (7).
MRI provides multiplanar imaging abilities (Fig. 7) and has better delineation of tumor and soft tissue interface than CT (Fig. 8). MRI is able to detect bone marrow infiltration by tumors, while CT cannot detect this kind of infiltration unless there is associated bony erosion. MRI is also more sensitive at evaluating retropharyngeal and deep cervical nodal metastases (11). Despite all these advantages, MRI is not suited to evaluate details of bone erosion, and CT should be performed when the status of the base of the skull needs to be evaluated.
PET/CT Scan There is an increasing use of PET/CT scan in the diagnosis of malignancies in the head and neck region. It is useful in the nasopharynx region to differentiate inflammation from malignant tumors and in the detection of local recurrence after therapy. It is also helpful in the search for the primary at the skull base when the patient presents with a metastatic lymph node (12).
Diagnostic Imaging Cross-sectional imaging either by computed tomography (CT) or by magnetic resonance imaging (MRI) is invaluable in the investigation of pathologies in the nasopharynx. These imaging studies show the tumor in the nasopharynx, as well as any abnormal structural alterations in the nasopharynx
Figure 7 extension.
SURGICAL RESECTION OF TUMORS IN THE NASOPHARYNX For NPC, which is chemoradiosensitive, concomitant chemoradiation is the primary treatment modality (13). For
(Left) Direct coronal MR showing a schwannoma extending through the skull base. (Right) Axial MR of the same patient showing intracranial
Chapter 17: Tumors of the Nasopharynx
271
Figure 9 Schematic CT, (Left) The dotted lines mark the osteotomies. (Right) The right maxilla attached to the anterior cheek flap is swung laterally to expose tumor in the nasopharynx (arrow).
Figure 8 Axial MR showing a chordoma (Ch).
those patients who have residual disease or develop recurrence in the nasopharynx after chemoradiation, surgical salvage has been shown to be effective (14). As the nasopharynx and its nearby regions are located in the central portion of the head, it is difficult to approach and to expose the region adequately to carry out an oncological resection. A few approaches have been reported over the years to expose the nasopharynx to allow an oncological resection to be carried out for tumors situated in the region. To approach the region from the superior or the posterior aspects is not practical, as the brain and the spinal column are in the way which limits surgical exposure. The anterior approach to the nasopharynx through the transantral route or the midfacial degloving procedures exposes the anterior part of the nasopharynx. This approach does not provide adequate exposure of the whole nasopharynx, including the superior and lateral walls. These anterior approaches, even combined with the controlled fracture of the hard palate followed by its downward displacement, only increase the exposure of the superior and posterior walls of the nasopharynx and not its lateral walls which are frequently involved by most tumors. The nasopharynx and the central skull base can be approached from the inferior aspect employing the transpalatal, transmaxillary, and transcervical approaches (15,16). This approach is useful for tumors located in the central and lower posterior wall of the nasopharynx. For more extensive tumors, especially those situated on the roof and on the lateral wall, the exposure of the lateral aspect of the tumor is not adequate. The dissection of the paranasopharyngeal space is difficult with this inferior approach. The internal carotid artery lying in this area has to be protected during the removal of tumor and with this approach the control of the vessel is suboptimal. The surgical approach to the nasopharynx and its vicinity from the lateral aspect through the infratemporal fossa has been described (17). With this approach, a radical mastoidectomy has to be performed and some important structures have to be mobilized, which include the internal carotid artery, the fifth cranial nerve, and the floor of the middle cranial fossa. The resultant morbidities following this approach are significant; in addition, considerable surgical expertise is required to employ this approach to the central skull base. This approach exposes directly the lateral wall of the na-
sopharynx, including the ipsilateral internal carotid artery. It, however, does not provide adequate access to the lateral wall of the nasopharynx on the opposite side and thus have limitations when the tumor in the nasopharynx crosses the midline. The nasopharynx can also be approached from the anterolateral aspect through the lateral swing of the hard palate and the maxillary antrum. The maxilla remains attached to the anterior cheek flap, where it derives its blood supply. This maxillary swing approach exposes the nasopharynx and its vicinity widely for an oncological resection (Fig. 9). The approach offers unobstructed view of the region from the anterior, inferior, and lateral aspects. With the extensive exposure of the region, the internal carotid artery lying outside the pharyngobasilar fascia can be palpated and closely observed during the resection of the tumor. Following extirpation of tumor, the whole osteocutaneous complex can be returned to its original position and fixed to the rest of the facial skeleton. We have employed this approach for resection of tumors in the nasopharynx and those extending to nearby regions. The outcome has been satisfactory with no mortality and minimal morbidity (18–20).
Preoperative Preparation The general condition of the patient should be evaluated to ensure that the patient is fit for the surgery. This includes the cardiopulmonary status and hematologic status. The patient’s blood should be typed and screened in case there is excessive blood loss, blood transfusion can be given, although this is seldom required. Any trismus should be eliminated as much as possible through mechanical stretching before the operation. The interalveolar distance should be more than 3 cm to allow the intraoral procedures, such as lifting the palatal flaps and the insertion of the curved osteotome to separate the pterygoid plates from maxillary tuberosity. Preoperatively, the orthodontist should take an impression of the upper alveolus of the patient in order to make a dental plate. This dental plate conforming to the contour of the whole upper alveolus facilitates the accurate placement of the swung maxilla after its return to the original position following the extirpation of tumor in the nasopharynx. While attaching to the opposite upper alveolus, the dental plate also contributes to immobilize the swung maxilla to facilitate healing. As the surgical field involves the oral cavity, intravenous broad spectrum antibiotics are started preoperatively.
272
Wei and Lam
SURGICAL TECHNIQUE Patient Position The operation is carried out under general anesthesia with a temporary tracheostomy performed for ventilation. This will leave the oral and nasal cavities free for the various operative maneuvers. The patient lies in supine position with the head slightly elevated. A temporary tarsorrhaphy suture is placed on the eyelid of the operative side while the other eye is covered with protective tape. The operative site is cleaned with antiseptic solution from the forehead down to the root of the neck and draped accordingly. The neck on the side of the operation is also prepared in case of the need for internal carotid artery control during the operation. The oral cavity is cleaned with antiseptic solution and a pharyngeal pack is inserted.
Incision The Weber-Ferguson incision is used. This starts at about 2 cm lateral to the outer canthus of the eye and should be close to the lower border of the zygomatic arch. The incision then extends medially at 3 mm below and parallel to the lower eyelashes. If the incision is placed too close to the eyelashes, edema may develop in the region above the incision, and if placed too far away, the patient may develop ectropion. When the incision reaches the medial canthus, it is curved inferiorly along the lateral border of the nose to reach the alar groove. The curve of the incision around the medial canthus is fashioned to be as obtuse as possible to ensure the tip of the skin has adequate blood supply to avoid necrosis. This is particularly important if the patient had previous radiotherapy. The incision continues inferiorly along the side of the nose to reach the midline, where it then turns downward at right angle, extends along the philtrum to divide the upper lip up to the vermilion. The incision over the vermilion part of the upper lip should be in a zigzag line to avoid subsequent contracture and upward retraction of the lip. This incision continues over the gum to reach the groove between the two central incisors (Fig. 10). The palatal incision is placed on the lingual aspect of the upper alveolus. It starts from the mucosa 0.5 cm from the canine on the opposite side and curves along the inner border of the upper alveolus toward the operative side, keeping a mucosal ridge of 0.5 cm. On reaching the third molar, it turns laterally toward the buccal mucosa behind the maxillary tuberosity (Fig. 11).
Figure 10 Facial incision for the maxillary swing approach to central skull base is marked.
The greater palatine vessel on the side of the swing should be divided between clamps. Following this, the elevated palatal flap can be lifted to expose the midline of the hard palate and posteriorly the attachment of the soft palate to the hard palate. This attachment should be divided to enter the posterior nasal cavity, thus detaching the soft palate from the hard palate.
Bony Dissection and Osteotomies A total of four osteotomies are required to detach the maxilla from the rest of the facial skeleton. The first horizontal osteotomy is placed on the anterior wall of the maxilla below the inferior rim of the orbit, thus the orbital floor is not disturbed. This osteotomy is extended laterally onto the zygomatic arch and then turns inferior, thus dividing only
Soft Tissue Dissection The transverse facial incision goes down onto the anterior wall of the maxilla. The soft tissue is elevated to expose only a strip of the anterior bony wall of the maxilla and the zygomatic arch in order to apply the oscillating saw and to place the miniplates. The infraorbital nerve and vessel should be divided and the anterior wall osteotomy usually goes through this infraorbital foramen (Fig. 12). The transverse incision continues to the lateral rhinotomy incision, which goes directly into the nasal cavity so that the entire upper and lower lateral cartilages on the side of the swing are lifted and turned medially up to the midline. The soft tissue over the incisor fossa is elevated to the anterior nasal spine in the midline. The amount of soft tissue lifted should again be limited to allow the placement of the miniplates and the insertion of the oscillating saw. The incision over the mucosa on the hard palate should be down to bone and the mucoperiosteum covering of the hard palate should be lifted from the canine tooth on the opposite side to the last molar tooth on the side of the swing.
Figure 11 The palatal incision is marked.
Chapter 17: Tumors of the Nasopharynx
Figure 12 The osteotomy line on the anterior wall of the maxilla is marked (arrow). Limited amount of soft tissue over the anterior wall of the maxilla is elevated, only adequate to expose a narrow strip of bone for the oscillating saw.
the lower half of the zygomatic arch. The oscillating saw then goes through the maxillary antrum to separate the posterior wall of the maxilla from its superior bony attachments. The second osteotomy divides the entire hard palate in the midline. Care should be taken to protect the elevated palatal flap and the soft palate from injury by the oscillating saw, both should remain intact after the osteotomy. Before making these two osteotomies, the miniplates intended for fixation of the maxilla, with holes for the screws drilled, should be placed across the proposed osteotomy paths. This will facilitate the repositioning of the maxilla after resection of the tumor at the skull base. In general, two 4-hole miniplates are employed: one across the osteotomy on the zygoma and the other across the midline below the nasal spine. The third osteotomy divides the lateral wall of the nose. It is placed above the upper border of the inferior turbinate and goes from anterior to posterior dividing the bony lateral wall of the nose together with the lacrimal duct. The fourth osteotomy separates the maxillary tuberosity from the pterygoid plates through the use of a curved osteotome (Fig. 13).
273
Figure 13 Palatal flap (P) retracted medially and a curved osteotome (arrow) is inserted to separate the pterygoid plates from the maxillary tuberosity.
prevertebral muscle can be removed together with the tumor to increase the resection margin and for similar reason; the anterior wall of the sphenoid sinus is removed with a drill. The sphenoid sinus is thus opened while it is not necessary to remove its mucosa. The wide exposure achieved after the maxilla is swung laterally allows the dissection of the paranasopharyngeal space under direct vision. Thus lymph node or the tumor which has extended to this space such as schwannoma, angiofibroma, etc. can be removed under direct vision. The position of the internal carotid artery lying outside the pharyngobasilar fascia can be identified by palpation during the dissection and thus protected from injury.
Reconstruction While the maxilla is swung laterally, the inferior turbinate is removed (Fig. 16) and an inferior antrostomy is performed to improve the drainage from the maxillary sinus. The mucosa over the removed inferior turbinate is spread out as a sheet and placed on the raw area in the nasopharynx as a free mucosal graft (Fig. 16). After resection of tumor localized in
Following the osteotomies, the maxilla with half of the hard palate drops inferiorly but is still attached to the anterior cheek flap. This whole osteocutaneous flap can be swung laterally as one unit to expose the nasopharynx and the central skull base (Fig. 14).
Tumor Resection After the maxilla is swung laterally, pterygoid plates are separated from the muscular attachments and removed. The pterygoid muscles are retracted after plication to stop bleeding from the venous plexus. The posterior part of the nasal septum, the vomer and septal cartilage, is removed to improve exposure of the opposite nasopharynx. Then the entire nasopharynx and the paranasopharyngeal space on the side of the swing are exposed to allow an oncological resection. Tumor located in the nasopharynx and its vicinity can be removed under direct vision. For recurrent NPC, it is frequently attached to the auditory tympanic tube. As such, the cartilaginous portion of this tube together with the lateral wall of the nasopharynx including the fossa of Rosenmuller on the side of the swing is removed with the tumor en bloc (Fig. 15). The
Figure 14 The maxilla attached to the anterior cheek flap is swung laterally as an osteocutaneous flap to expose the nasopharynx and central skull base (N).
274
Wei and Lam
Figure 16 (Left) Inferior turbinectomy carried out on the laterally swung maxilla. (Right) The mucosa from the inferior turbinate as a free graft (G) is laid onto the raw area in the nasopharynx following tumor resection.
drilled on the skull base bone to facilitate the insertion of sutures. The vascular pedicle is transposed inferiorly lateral to the wall of the oropharynx and the ascending ramus of the mandible. The microvascular vascular anastomosis is carried out with the facial vessels at the lower border of the mandible or in the neck.
Closure
Figure 15 Nasopharyngectomy specimen showing the tumor (T) and the Eustachian tube crura; the Eustachian tube opening is marked by a plastic tube (arrow).
the nasopharynx, the swung maxilla is returned to its original position and it is fixed to the rest of the facial skeleton with the two miniplates. One additional wire loop is used to approximate the upper medial angle of the swung maxilla and the nasal bone. This three-point fixation together with the prefabricated dental plate gives a stable reconstruction of the upper alveolus (Fig. 17). If a significant resection of the soft tissue such as the soft palate or lateral pharyngeal wall is done for oncological clearance, additional reconstruction with free tissue transfer is required. This also applies when a significant size of skull base bone is removed with or without resection of the dura, particularly in patients with previous radiotherapy. In addition to filling the soft tissue defect, the free tissue also contributes increased vascularity to prevent the development of osteoradionecrosis. The free flap of choice is the rectus abdominis muscle with or without the overlying skin island. When there is a pharyngeal defect, the skin island is useful for the reconstruction of this defect. Otherwise, the muscle alone is adequate to fill the soft tissue defect. The muscle component is insetted to fill the defect in the nasopharynx through stitching to the pterygoid muscle and tissue at the skull base. Occasionally, holes have to be
After returning and fixing of the swung maxilla to the facial skeleton with the two miniplates and one wire loop, the hard palate mucoperiosteum is repositioned (Fig. 18). Only a few stitches are required near the maxillary tuberosity. Stitching the palatal flap along the entire alveolar margin might produce tenting and deprive of its contact with the hard palate (21). The insertion of the prefabricated dental plate will ensure a close contact of this flap with the underlying bone. A nasogastric tube is inserted via the opposite nostril for postoperative feeding. The posterior part of the nose on the side of the swing might be packed with a Foley catheter.
Figure 17 The maxilla is fixed to the rest of the facial skeleton with miniplates (arrow heads) and one wire loop at the upper medial angle (arrow) of the maxilla.
Chapter 17: Tumors of the Nasopharynx
275
on the third postoperative day. At about 1 week after the operation, oral feeding can be started and nasal pack with the Foley catheter can also be removed. For those patients who had previous radiotherapy, oral feeding is usually started about 2 weeks after the operation.
Complications and Avoidance
Figure 18 (Left) The hard palate mucoperiosteum is repositioned (M) and only two stitches (arrow) are placed near the maxillary tuberosity, (Right) The prefabricated dental plate is placed to maintain the position of the palatal flap.
The blown-up balloon of the catheter will keep the mucosal graft in position and reduce the amount of blood or serum dripping into the oropharynx. The anterior nasal cavity is packed with Merocel and the facial wound is closed with interrupted suture in layers (Fig. 19).
Postoperative Care The patient is advised to have his head elevated at 30 degrees to reduce facial edema and nasal gastric tube feeding can be started after the patient recovers from general anesthesia. The tracheostomy tube can be removed when the patient has regained the ability to cough and this can usually be done
For patients who had previous radiotherapy, the possibility of development of osteoradionecrosis of the skull base bone is high. Yet with limited bone exposure, conservative management is usually adequate to limit such complications. For extensive area of osteoradionecrosis, the infection might extend intracranially. Early recognition of the problem and extensive adequate debridement followed by reconstruction with a microvascular free flap is essential for salvage. To avoid the development of osteoradionecrosis, all the exposed bone should be covered with soft tissue and when necessary, a microvascular free flap should be done initially. When the maxillary swing procedure was initially reported, the split of the mucoperiosteum and the hard plate was in the same plane. Thus the development of palatal fistula used to be one of the main complications. With the modification of the incision over the palate and the elevation of the palatal flap, the osteotomy of the hard palate and the incision of the mucoperiosteum are no longer in the same plane. Thus palatal fistula can be avoided in most patients nowadays. Another complication is the development of trismus. This is frequently encountered in patients who had previous radiotherapy. The trismus responds to passive mechanical stretching. Once interalveolar distance reaches 2 cm, normal oral feeding is usually smooth. Occasionally, patients may develop ectropion and this is related to the contracture of the lower eyelid. Revision of the scar together with the insertion of a small piece of full thickness skin graft will solve the problem.
Follow-up and Rehabilitation During follow-up sessions, endoscopic examination of the nasal cavity and the nasopharynx together with irrigation and cleansing should be carried out to reduce the incidence of infection. This is particularly important for patients with previous radiotherapy. The dental plate is removed at 1 week after operation and the patient is advised to have soft diet for 1 month to allow bony reunion of the osteotomies. All patients are advised to have active and passive stretching exercises to reduce trismus. Any recurrence of pathology in the nasopharyngeal region can be detected with endoscopic examination and other imaging studies, which are carried out as indicated.
REFERENCES
Figure 19 At the completion of operation, the facial wound is closed and the nasogastric tube inserted.
1. Harrison DFN. The natural history, pathogenesis, and treatment of juvenile angiofibroma. Personal experience with 44 patients. Arch Otolaryngol Head Neck Surg. 1987;13:936–942. 2. Nicholls JM. Nasopharyngeal carcinoma:Classification and histological appearances. Adv Anat Path. 1997;4:71–84. 3. Sham JS, Wei WI, Lau SK, et al. Serous otitis media. An opportunity for early recognition of nasopharyngeal carcinoma. Arch Otolaryngol Head Neck Surg. 1992;118;794–797. 4. Wei WI, Sham JS, Zong YS, et al. The efficacy of fiberoptic endoscopic examination and biopsy in the detection of early nasopharyngeal carcinoma. Cancer. 1991;67:3127–3130. 5. Chien YC, Chen JY, Liu MY, et al. Serologic markers of EpsteinBarr virus infection and nasopharyngeal carcinoma in Taiwanese men. N Engl J Med. 2001;345:1877–1882.
276
Wei and Lam
6. Lo YM, Chan LY, Lo KW, et al. Quantitative analysis of cell-free Epstein-Barr virus DNA in plasma of patients with nasopharyngeal carcinoma. Cancer Res. 1999;59:1188–1191. 7. Wei WI, Yuen AP, Ng RW, et al. Quantitative analysis of plasma cell-free Epstein-Barr virus DNA in nasopharyngeal carcinoma after salvage nasopharyngectomy: A prospective study. Head Neck. 2004;26:878–883. 8. Sham JS, Cheung YK, Choy D, et al. Nasopharyngeal carcinoma: CT evaluation of patterns of tumor spread. AJNR Am J Neuroradiol. 1991;12:265–270. 9. Chong VF, Fan YF, Khoo JB. Nasopharyngeal carcinoma with intracranial spread: CT and MR characteristics. J Comput Assist Tomogr. 1996;20:563–569. 10. Fang FM, Leung SW, Wang CJ, et al. Computed tomography findings of bony regeneration after radiotherapy for nasopharyngeal carcinoma with skull base destruction: Implications for local control. Int J Radiat Oncol Biol Phys. 1999;44:305–309. 11. Dillon WP, Mills CM, Kjos B, et al. Magnetic resonance imaging of the nasopharynx. Radiology 1984;152:731–738. 12. Fukui MB, Blodgett TM, Snyderman CH, et al. Combined PETCT in the head and neck: Part 2. Diagnostic uses and pitfalls of oncologic imaging. Radiographics. 2005;25:913–930.
13. Wei WI, Sham JS. Nasopharyngeal carcinoma. Lancet. 2005;365:2041–2054. 14. Wei WI. Cancer of the nasopharynx: Functional surgical salvage. World J Surg. 2003;27:844–848. 15. Fee WE Jr, Roberson JB Jr, Goffinet DR. Long-term survival after surgical resection for recurrent nasopharyngeal cancer after radiotherapy failure. Arch Otolaryngol Head Neck Surg. 1991;117:1233–1236. 16. Morton RP, Liavaag PG, McLean M, et al. Transcervicomandibulo-palatal approach for surgical salvage of recurrent nasopharyngeal cancer. Head Neck. 1996;18:352–358. 17. Fisch U. The infratemporal fossa approach for nasopharyngeal tumors. Laryngoscope. 1983;93:36–44. 18. Wei WI, Lam KH, Sham JS. New approach to the nasopharynx: The maxillary swing approach. Head Neck. 1991;13:200–207. 19. Wei WI, Ho CM, Yuen PW, et al. Maxillary swing approach for resection of tumors in and around the nasopharynx. Arch Otolaryngol Head Neck Surg. 1995;121:638–642. 20. Wei WI. Nasopharyngeal cancer: Current status of management. Arch Otolaryngol Head Neck Surg. 2001;127:766–769. 21. Ng RW, Wei WI. Elimination of palatal fistula after the maxillary swing procedure. Head Neck. 2005;27:608–612.
18 Clival Tumors Franco DeMonte, Mark J. Dannenbaum, and Ehab Y. Hanna
INTRODUCTION
Multiplanar MRI with and without contrast enhancement best demonstrates the extent of the tumor and best identifies important adjacent neurovascular structures such as the brainstem, optic nerves and chiasm, and internal carotid and basilar arteries. Fat suppression techniques highlight tumoral contrast enhancement within or adjacent to fatty areas such as the orbits and bone marrow (5,6). Cerebral angiography is not typically necessary for the diagnosis of clival tumors. It is used when a detailed view of the head and neck and intracranial vasculature is desired. Important anatomic variations are noted, such as the pattern of venous drainage or the integrity of the circle of Willis. Vascular distortion or narrowing is well identified. Preoperative temporary balloon occlusion testing of the ICA may be performed if ICA occlusion or sacrifice is planned or if risk of arterial injury is deemed high. One or more complimentary investigations of cerebrovascular reserve such as transcranial Doppler (TCD), cerebral blood flow (CBF), or single photon emission computed tomography (SPECT) may be used in conjunction with the temporary balloon occlusion test. The differential diagnosis of an invasive clival mass includes chordoma, chondrosarcoma (although these tumors are usually found in a paramedian location), pituitary adenoma, metastasis, meningioma, nasopharyngeal carcinoma, and primary sphenoid sinus malignancy (Table 1). On conventional spin echo T1-weighted MRI, chordoma produces an intermediate to low signal intensity and is easily recognized among the high signal intensity of the fat within the clivus. Occasionally, small foci of hyperintensity may be seen within the tumor on T1-weighted sequences. This finding may represent either intratumoral hemorrhage or mucin production by the tumor. T2-weighted MRI usually demonstrates high signal as a result of the high-fluid content of the vacuolated cellular components. Areas within the tumor containing calcium, hemorrhage or densely proteinaceous mucus will most often produce T2 hypointensity. Interlaced low-intensity septations separating high signal intensity lobules are commonly visualized and correspond with the multilobular gross morphology of the tumor. Contrast enhancement is seen in the majority of chordomas and ranges from moderate to marked. The enhancement pattern of the tumor is often described as a “honeycomb” appearance that is created by intratumoral areas of low signal intensity [Fig. 2(A)–2(C)]. Fat suppressed images are useful for differentiating enhanced tumor margins from adjacent bright fatty bone marrow (5,6). As opposed to intracranial chordomas, which have a predilection for the midline of the skull base, the majority of chondrosarcomas are located more laterally along the petroclival fissure. Chondrosarcomas may, however, have a midline origin that makes preoperative differentiation from chordoma more challenging as both have similar signal intensity
This chapter focuses on the surgical treatment of clival tumors. Since clival chordoma is the archetypal tumor in this location, specific comments regarding this tumor type will be made. Irrespective of the pathology, the bone of the clivus may be infiltrated by tumor and require resection beyond the grossly visible tumor margin. The majority of tumors in this location are large at diagnosis and extend to the anterior cranial fossa or parasellar region or ventrally into the nasal cavity, paranasal sinuses or nasopharynx or to the middle fossa.
SURGICAL ANATOMY The clivus is formed from the part of the basilar occipital bone that extends anteriorly and superiorly from the foramen magnum and from the body of the sphenoid bone. These two bones articulate at the spheno–occipital synchondrosis. The clivus thus extends from the dorsum sella and the posterior clinoid processes to the foramen magnum. At its superior aspect, the clivus is bordered laterally by the petroclival fissure (synchondrosis) and the petrous temporal bone. The petroclival fissure terminates in the jugular foramen posterolaterally. The pharyngeal tubercle located on the anterior and inferior surface of the clivus is the site of attachment of the midline pharyngeal raphe. A rich venous plexus exists between the periosteal and meningeal layers of the clival dura. The inferior petrosal sinus runs along the petroclival fissure and connects the clival venous plexus and the posterior cavernous sinus to the jugular bulb. The medulla, pons, and midbrain and the vertebrobasilar arterial tree lay directly behind the clivus. The abducens nerve penetrates the meningeal layer of the clival dura at Dorello’s canal and courses superiorly, medial to the petrous apex, to enter the posterior cavernous sinus. The posterolateral aspects of the basiocciput form the occipital condyles and transmit the hypoglossal nerves (1–4).
DIAGNOSTIC IMAGING, REGIONAL PATHOLOGY, AND DIFFERENTIAL DIAGNOSIS High-resolution computed tomographic (CT) scanning allows the accurate assessment of bony destruction by the tumor, which is universally present in chordomas (5) (Fig. 1). Integrity of the optic and carotid canals can be best assessed in this way. Tumoral calcification is well seen if present. If there is a question of occipitocervical instability, plain x-ray films of the cervical spine, both in flexion and extension, are performed. 277
278
DeMonte et al. Table 1
Differential Diagnosis of Clival Tumors
Chordoma Chondrosarcoma Nasopharyngeal carcinoma Pituitary macroadenoma Fibrous dysplasia Metastasis Meningioma Rhabdomyosarcoma Craniopharyngioma Dermoid Epidermoid
Figure 1 Axial CT scan with bone windowing in patient with clival chordoma. Note the destruction of the central portion of the clivus with a loss of the cortical bone and islands of calcification within the tumor. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
on T1- and T2-weighted MR images [Fig. 3(A)–3(C)]. Chondrosarcoma can sometimes be distinguished by the presence of linear, globular or arc-like matrix calcifications. Clival meningiomas represent another, yet, small subset of the pathologies encountered in this area. On MR imaging they appear as a well circumscribed, avidly and homogenously contrast enhancing mass with a dural attachment. They do not have the bony destruction associated with chordomas, but may be associated with hyperostosis. Nasopharyngeal cancers including nasopharyngeal carcinoma, lymphoma, and plasmacytomas must always be considered in the differential diagnosis of clival chordoma. They usually extend more anteriorly and may be associated with head and neck lymphadenopathy. These masses may produce lytic destruction of the skull base as well, which when centrally located may produce imaging that exactly mimics clival chordoma.
Pituitary macroadenomas may at times appear to involve and/or originate from the clivus. The inability to identify the pituitary gland in the presence of a large clival mass should raise the possibility of pituitary macroadenoma [Fig. 4(A) and 4(B)]. A markedly elevated serum prolactin level may obviate the need for biopsy or resection. Rhabdomyosarcoma represents another pathology that in children may develop in the clival region. This lesion takes origin in the nasopharynx and manifests as a large, bulky, intra- and extracranial tumor with associated lytic osseous destruction of the skull base.
CLINICAL ASSESSMENT All patients have a complete history and physical examination including a neuro–ophthalmologic evaluation, which assesses visual acuity and fields, pupillary function, and ocular motility. Less than one-third of these patients have a normal neuro–ophthalmologic examination. Patients with upper clival tumors undergo a complete endocrinologic assessment to identify evidence of pituitary or hypothalamic dysfunction. Patients with mid to lower clival tumors are assessed by an otolaryngologist with specific attention to hearing and lower cranial nerve function. An audiogram and direct laryngoscopy are usually performed.
Figure 2 (A) Axial T1-weighted MRI, (B) axial T2-weighted MRI, (C) axial T1-weighted postcontrast MRI of the skull base in a patient with a left lower clival chordoma with extension to the cervicomedullary junction. On precontrast T1-weighted imaging, note the loss of the marrow fat signal in the lower left clivus. The tumor is isointense to brain on T1-weighted imaging. On T2-weighted imaging, the tumor is bright but there are areas of heterogeneity within the tumor of lesser signal intensity. Following contrast administration, only mild contrast enhancement is identified. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
Chapter 18: Clival Tumors
279
Figure 3 (A) Axial T1-weighted MRI, (B) axial T2-weighted MRI, (C) postcontrast T1-weighted MRI of a patient with a right skull base chondrosarcoma. On T1-weighted imaging, the tumor is of decreased signal intensity. Note the loss of marrow fat in the right petrous apex. On T2-weighted imaging, the lesion is homogenously of high signal, which is slightly unusual for chondrosarcoma that tends to have more of a heterogenous T2-signal, similar to chordoma. There is moderate to marked homogenous enhancement following contrast administration. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
The symptoms resulting from clival tumors depend on the specific sites of their extension Table 2. The most common manifestation is a headache located in the occipital or occipitocervical area and may be aggravated by changes in neck position. Diplopia is the initial symptom experienced by most patients and is almost always the result of abducens neuropathy. Larger tumors may produce additional cranial nerve palsies manifesting as decreased visual acuity, facial weakness or numbness, hearing loss, dysphagia, and dysphonia. When tumors enlarge and produce brainstem or cerebellar compression, patients may develop ataxia, dysmetria, and motor weakness. When tumor extends into the retropharyngeal space or the nasal cavity, symptoms such as nasal obstruction, epistaxis, or throat fullness may develop.
SURGICAL MANAGEMENT There have been numerous surgical approaches used to access the clivus. The choice of one approach over another needs to factor in the parameters of tumor location, size, extension, unique patient anatomy and functional requirements, and experience, expertise, and preference of the surgical team. Small, midline upper clival tumors without lateral extension are, for example, probably best removed through a transsphenoidal approach (open or endoscopic), while large tumors with lateral and intradural extensions may require both a transfacial and intracranial approach for adequate tumor removal. Table 3 lists the various surgical approaches, which are useful for clival tumor resection. Although each will be discussed briefly, the transmaxillary, transmandibular, transbasal, and transcondylar approaches will be dealt with in greater detail.
Transsphenoidal Approach The transsphenoidal approach works well for smaller tumors located in the upper and middle clivus. Excellent exposure of the sphenoid sinus, sella turcica, and upper and middle clivus is possible (7). It has the disadvantages of limited inferior and lateral exposure although the lateral exposure can be improved by entry into the medial maxillary sinus and removal of the pterygoid plate. The surgical field is typically deep and narrow and thus there is little room to allow aggressive bone resection (8). These disadvantages are much less of an issue with current endoscopic instrumentation, although Figure 4 (A) Sagittal postcontrast MRI imaging, (B) coronal postcontrast MRI imaging of a patient with a large clival mass scheduled for open surgical resection. It was noted that the pituitary gland could not be visualized clearly on any of the sequences. A serum prolactin level was performed and was noted to be massively elevated at 345,000 ng/mL. The tumor responded very well to medical treatment with cabergoline without biopsy or resection. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
Table 2
Symptoms of Clival Tumors Based on Site of Origin
Upper clivus Mid clivus
Lower clivus
Pituitary endocrinopathy, visual loss, chiasmal syndrome, cavernous sinus syndrome Nasopharyngeal mass, abducens nerve palsy, multiple cranial neuropathies, brainstem syndrome, hydrocephalus, cerebellopontine angle syndrome Hypoglossal nerve palsy, foramen magnum syndrome
280
DeMonte et al.
Table 3 Surgical Approaches to the Clivus Transsphenoidal-–open, endoscopic Transsphenoethmoidal Transoral-transpalatal Transmaxillary Transmandibular circumglossal retropharyngeal Extended transbasal Anterior transtemporal Posterior transtemporal Transcondylar
further refinement is necessary (1, 2). Entirely endoscopic transsphenoidal resection of clival tumors is increasingly being reported [Fig. 5(A) and 5(B)]. (See chapters 7 and 23)
Transsphenoethmoidal Approach In order to expand the exposure offered by the transsphenoidal approach, an external ethmoidectomy with or without a medial maxillectomy can be employed. Lalwani et al. found this approach to be “adequate for the majority of tumors and disease processes present in the sphenoid sinus and clivus” (9). The addition of the medial maxillectomy in instances of a narrow ethmoid sinus or inferiorly extending tumor allows improved access. Disadvantages include the necessity of a facial incision, and the relatively limited exposure. The inferior limits can be extended by a medial maxillectomy as described above, but lateral reach is limited to approximately 2 cm from the midline.
Transoral–Transpalatal Approach Following the placement of an appropriate oral retractor system, the posterior pharyngeal wall and soft palate are divided to expose the clivus and upper cervical spine. Most commonly employed for odontoidectomy, the exposure obtained with this approach is relatively quite limited (10). It is adequate for tumor biopsy or in the case of small lower clival tumors
[Fig. 6(A)–6(C)]. Repair of intraoperative cerebrospinal fluid leakage is problematic. Although fat and/or fascial tissue can be placed posterior to the pharyngeal wall, the pharyngeal wall rarely approximates well and gaps are common, thus increasing the risk of CSF leakage. The transoral–transpalatal approach should not be the primary approach to intradural lesions at the foramen magnum unless other approaches prove ineffective (11). (See chapter 28)
Transmaxillary Approaches The transmaxillary approaches may prove useful for clival tumors that extend into the nasopharynx or craniocervical junction with minimal lateral extension [Fig. 7(A) and 7(B)]. Numerous variations of this approach have been described. Most are based on a Le Fort I osteotomy with or without midline splitting of the hard and soft palates or a unilateral maxillotomy with median or paramedian splitting of the hard and soft palates (12–15). We prefer approaches that avoid displacing or sectioning the palate and alveolar arch (3,16– 18). Access to the maxilla is obtained via a facial degloving approach or through a lateral rhinotomy with lip split (3,19). Lateral exposure is limited by the pterygoid plates, the ICA at the level of the foramen lacerum and cavernous sinus, the hypoglossal canal and the jugular foramen. Removal of the pterygoid plates allows for access to the ITF (20,21). Disadvantages of the transmaxillary approaches include the risk of ischemic osteonecrosis with multisegment osteotomies. This risk can be minimized by subtotally splitting the soft palate if palatal osteotomies are necessary and by leaving wide soft tissue attachments to freed maxillary compartments during swing procedures, or avoided entirely by transantral approaches. The principal disadvantage of the transmaxillary approach, however, is the inability to reliably stop CSF leakage. Direct repair of the dura is rarely possible and must rely on packing with fat, fascia, and fibrin glue without the availability of firm tissue support to hold the repair in place. Elevation and use of vascularized mucosal flaps (septal, nasal, or turbinate) helps with healing and has reduced the CSF leak rate.
Positioning The patient is positioned supine on the operating table with the head slightly elevated, extended, and rigidly fixated. Intraoperative image guidance is particularly useful in this approach. Best visualization is obtained by standing on the side opposite the maximum extent of the tumor. For example a clival tumor extending posteriorly and to the left is best approach through the right maxilla.
Incision and Soft Tissue Dissection
Figure 5 (A) Axial T2-weighted MRI, (B) sagittal postcontrast T1-weighted MRI identify a lesion of the posterior nasal cavity, sphenoid and ethmoid sinuses, and clivus with minimal lateral extension. This tumor was accessed and resected with an entirely endoscopic technique. A chronic dural perforation was identified and repaired. Despite this primary repair, postoperative CSF leakage was noted. The patient was taken back to the operating room where a repeat endoscopic procedure was performed. The dural defect was widened and an intra- and extradural repair was performed with an “onlay” technique with fascia and fat, which resulted in resolution of the spinal fluid leakage. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
The upper lip is retracted and a sublabial incision is made along the gingival sulcus extending to the lateral maxillary buttress. The soft tissues of the face are then elevated subperiosteally to expose the piriform aperture, infraorbital foramen, and floor of the nasal cavity.
Bony Dissection, Osteotomies, and Tumor Resection A complete anterior maxillary antrotomy is performed preserving the infraorbital nerve. The medial wall of the antrum, the palatine bone, and the posterior wall of the antrum are removed exposing the pterygopalatine/pterygomaxillary fossa. The sphenopalatine artery is identified and coagulated if necessary. The frontal process of the maxilla is divided as close to the alveolar ridge as possible and as superiorly as possible (Fig. 8). The nasal mucosa is opened and the turbinates
Chapter 18: Clival Tumors
281
Figure 6 (A) Sagittal postcontrast T1-weighted MRI reveals a small enhancing lesion in the lower clivus. Biopsy confirmed it as a chordoma. (B) Intraoperative photograph of a transoral approach and resection of this small clival lesion. The oral retractor is in place and the pharyngeal mucosa and musculature is being retracted laterally. The tumor was resected in its entirety. (C) Intraoperative frameless stereotactic navigation was utilized and found to be helpful in the localization and resection of this lesion. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
dislocated anteriorly or removed. It is preferable to try to preserve the turbinates in order to keep the option of a turbinate flap. The nasal septum is dislocated to the opposite side and the vomer and sphenoid rostrum identified. The vomer is then removed facilitating entry into the sphenoid sinus. The entire anterior wall and floor of the sphenoid sinus is removed. The pharyngobasilar fascia and overlaying nasopharyngeal mucosa are elevated with cautery and flapped inferiorly. Using the operating microscope, the infraorbital nerve is followed to the pterygomaxillary fissure, which lies just behind the posterior wall of the maxillary sinus. The nerve is dissected free of the adipose tissue and followed to the foramen rotundum. The foramen is located at the top of the pterygoid plate, just lateral to the sphenoid sinus. The foramen is widened with a high-speed drill to expose the dura of the temporal fossa floor lateral to the foramen and the dura of the medial cavernous sinus medially. Removing the bone superomedial to the foramen rotundum unifies the foramen rotundum and superior orbital fissure [Fig. 9(A) and 9(B)]. Removal of the lateral wall of the sphenoid sinus connects this expo-
Figure 7 (A) Postcontrast T1-weighted MRI at the level of the petrous carotid, (B) postcontrast T1-weighted MRI at the level of the cavernous carotid. This patient presented with diplopia. Biopsy was consistent with a primary low-grade adenocarcinoma. The patient underwent resection via a left transmaxillary approach. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
sure to the midline exposure of the sellar region. The bone of the lateral sphenoid sinus is removed to expose the medial CS further. The clivus is thus fully exposed. Clivectomy is performed to the extent needed [Fig. 10(A) and 10(B)].
Reconstruction and Closure If the dura has been penetrated, dural closure is enhanced by placement of a fascial graft in addition to fat and fibrin glue. This repair is reinforced with vascularized nasal wall, floor, and septal flaps. A turbinate flap can also be used. The nasal septum is relocated and the frontal process of the maxilla is plated into its normal anatomic position. As with all of the anterior transfacial approaches, intermittent graduated lumbar spinal drainage may be used.
Figure 8 In this surgical photograph, the antrum of the left maxilla has been widely opened by a large anterior antrotomy. The maxillary antrum and medial maxilla have been unified with the nasal cavity by removal of the medial antral wall and of the medial maxilla. The septum has been displaced to the right and the vomer removed. The face of the sphenoid has been opened and surgical hemostatic agent is on the clival dura following resection of the primary adenocarcinoma identified in Figure 7(A) and 7(B). Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
282
DeMonte et al.
Figure 9 (A) Artist’s representation of the expansion of the transmaxillary approach to the anterior cavernous sinus and infratemporal fossa. Under the operating microscope, the foramen rotundum is opened both medially, unified with the sphenoid sinus, and laterally to expose the temporal dura. Continued removal of the pterygoid plate opens the foramen ovale and allows access to the infratemporal fossa. With complete removal of the pterygoid plate, the entire sphenoid sinus is accessed and the entire anterior wall and floor are removed. The pharyngobasilar fascia can be detached and thus the entire clivus is within the surgical field. (B) Intraoperative photograph shows the orbital tissues being protected by the brain retractor. The suction is in the sphenoid sinus, palpating the medial cavernous sinus wall. The foramen rotundum has been widened to expose the temporal dura lateral to it and superior to it. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
Transmandibular, Circumglossal, Retropharyngeal, Transpalatal Approach to Clivus and Upper Cervical Spine Initially described by Biller et al. (22) and popularized by Krespi (23–25) and later by Ammirati (26,27), the transmandibular approach to the skull base provides simultaneous exposure of the middle and lateral compartments of the skull base, allows excellent vascular control and access to cranial nerves IX–XII, and by straight forward expansion allows an exposure that extends from the ipsilateral ITF to the contralateral medial pterygoid plate, from the anterior cranial fossa to the lower clivus and the anterior cervical spine down to C7 (28). The main indications for this approach are large tumors involving both the middle and lateral compartments and tumors of the craniocervical junction and upper cervical spine. Ideally suited for extradural pathology, intradural extensions can be removed by this approach, although the risk of CSF leakage and meningitis increases (28). Even though dural repair is, as already mentioned, rarely water-
tight in this region, the elevation and subsequent replacement of the laterally based pharyngeal flap helps buttress any needed dural repairs. Along with the judicious use of subcutaneous fat and fascial grafts, fibrin glue, and lumbar spinal drainage, this minimizes the incidence of CSF leakage. A good deal of anatomic dissection is required for this approach, which results in predictable morbidity. A temporary tracheostomy is required. Conductive hearing loss and serous otitis media result from section of the Eustachian tube. Temporary swallowing difficulties induced by the circumglossal and palatal incisions, the extensive retropharyngeal dissection, and the section of the tensor and levator palatini muscles occasionally necessitate the insertion of a gastrostomy. Preoperative consultation with dentistry should be obtained to fashion a palatal stent to be used for support of the palatal mucosa after closure. This stent allows for improved apposition of the mucosa against the residual hard palate. This helps to prevent acute mucosal loss.
Chapter 18: Clival Tumors
Figure 10 (A and B) Postcontrast axial MRI imaging following resection of a primary adenocarcinoma via a left transmaxillary approach. Note complete removal of the tumor from the medial cavernous sinus wall and carotid artery. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
Due to this inherent morbidity, long and careful consideration should be given to the use of this approach. It should be reserved for instances where surgical cure or significant palliation are likely (Fig. 11) (28).
Positioning, Incision, and Soft Tissue and Bony Dissection With the patient in rigid cranial fixation in the supine position with the head slightly extended, a tracheostomy is performed (Fig. 12). A curvilinear incision is then begun just below the mastoid and extended inferiorly and medially along a skin crease to the mentum and through the lip split (Fig. 13). Subplatysmal flaps are elevated to expose the upper neck and submandibular gland and surrounding tissues. A selective neck dissection allows for the clear identification and preservation of the lingual and hypoglossal nerves. The ICA and IJV are identified and vessel loops are placed. Continued elevation of the flap exposes the mandible. The mandibulotomy site is carefully marked and an internal fixation plate is placed. The plate position is carefully marked,
283
Figure 12 Positioning for patient undergoing a transmandibular, circumglossal retropharyngeal approach to the clival chordoma illustrated in Figure 11. Use of intraoperative frameless stereotactic navigation aids in surgical planning and execution. The incision begins at the tip of the mastoid, curves forward to the midline and curves around the chin and then terminates in a midline lip split. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
and the plate removed and saved for closure. A stair-step mandibulotomy is performed between the lower medial two incisors (Fig. 13). Dissection follows the floor of the mouth posteriorly towards the glossopharyngeal sulcus. This allows the mandible to swing laterally and the tongue medially. The styloid process is palpated and the muscular attachments are separated. The ICA, IJV, and CNs IX–XII are traced superiorly to the skull base. As the incision approaches the anterior tonsillar pillar, it splits into two limbs. The upper limb extends to the soft palate, which is separated from its lateral attachments (Fig. 14). This incision is then carried onto the hard palate approximately 1 cm medial to the alveolar ridge. It then passes anteriorly around to the contralateral hard palate. The base of the hard palate is resected as needed for
Figure 11 Sagittal postcontrast T1-weighted MR imaging of a patient with a large and complex clival chordoma. Panel A reveals the tumor prior to any surgical intervention. Panel B reveals residual tumor following first stage transcondylar resection of the portion of the tumor, which was causing marked neural compression. Panel C reveals the final appearance following transmandibular, circumglossal retropharyngeal resection of the anterior components of the tumor. It is complex tumors, like this, that may require extensive and multi-staged procedures to maximize resection. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
284
DeMonte et al.
Figure 13 Artist’s representation of the incisions and stair-step mandibulotomy necessary for the transmandibular, circumglossal retropharyngeal approach to the clivus. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
visualization into the sphenoid sinus. The lower limb of the incision extends into the hypopharynx passing lateral to the tonsil and the orifice of the Eustachian tube. The Eustachian tube is transected and retropharyngeal dissection follows the subperiosteal plane to expose the clivus and upper cervical spine. This pharyngeal flap is elevated and rotated medially. The clivus and upper cervical spine are covered by the longus capitis muscles and the prevertebral fascia. Removal of the posterior hard palate reveals the posterior bony nasal septum (vomer) and the anteroinferior face of the sphenoid sinus (Fig. 15). Removal of this bone allows wide exposure of the sphenoid sinus and the sella turcica and upper clivus. At this point, clivectomy may be performed and may extend from the sella turcica to the foramen magnum (Fig. 16).
Figure 14 Artist’s representation of the transmandibular, circumglossal retropharyngeal approach to the clivus. In this illustration, the mandible has been reflected laterally keeping the lingual and hypoglossal nerves intact. The mucosal incisions at the level of the anterior tonsillar pillar extend superiorly onto the soft palate and subsequently onto the hard palate and inferiorly along the lateral pharynx in order to rotate the pharynx medially. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
The mandible is reapproximated and the reconstruction plate secured. Drains are placed in the neck and a nasogastric tube is inserted. Precise alignment of the vermilion border is assured during lip closure and the platysma and skin are closed in a standard fashion. Graduated, intermittent lumbar spinal drainage may be utilized if the dural closure is tenuous.
Reconstruction and Closure If the dura is deficient, repair with fascial and subcutaneous fat grafts and fibrin glue is performed. In this case, a layer of fascia is placed intradurally with its edges extending beyond the dural opening. If possible, this is tacked down with sutures. A second extradural fascial layer is placed and reinforced with fibrin glue and autologous fat. This can be reinforced with vascularized nasal septal flaps. Closure begins by reattaching the superior constrictor muscle to the muscles at the base of the skull. The palatal flap is reapproximated and the soft-palate and hard-palate mucosae are sutured. The preoperatively fashioned palatal stent is placed to support the palatal mucosa. The floor of the mouth is closed in a double layer. The pharyngeal flap is allowed to return to its normal position and closure is by careful reapproximation of the posterolateral mucosal edges.
Figure 15 Intraoperative photograph following completion of the approach phase of the transmandibular approach to the clivus. The mucosa of the palate is being retracted laterally and the hard palate has been resected in order to access the sphenoid sinus. The tumor is identified deep to the pharyngobasilar fascia and longus coli muscles. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
Chapter 18: Clival Tumors
285
teotomy (removal of supraorbital bar), ethmoidectomy, sphenoidectomy, and the extradural resection of the clivus. The limits of exposure are the foramen magnum inferiorly and the hypoglossal canals inferolaterally, while the intracavernous carotid arteries form the superolateral limits. The posterior clinoids, dorsum sella, and the region behind and above the pituitary gland are obscured by the gland, rendering this a blind area via the extended transbasal approach (Fig. 18).
Positioning, Incision, and Soft Tissue Dissection
Figure 16 Artist’s representation of the transmandibular approach to the clivus following clivectomy. The bone has been removed from the sella and from the undersides of the optic nerves. Bone removal can continue inferiorly down to the foramen magnum. At this point, it can be extended laterally to the hypoglossal canals bilaterally. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
Extended Transbasal Approach Developed by Tessier and Derome, the transbasal approach to the clivus has subsequently been modified to incorporate orbital osteotomies in an effort to reduce frontal lobe retraction and to allow exposure of more posteriorly and laterally related structures such as the medial walls of the cavernous sinus and the hypoglossal canals (Fig. 17) (12,29–34). The approach consists of a bifrontal craniotomy, bilateral orbital os-
The patient is placed in rigid cranial fixation in the supine position with the head slightly extended. Following a bicoronal scalp incision, the scalp is elevated sharply from the periosteum and loose connective tissue layers (collectively named the pericranium) immediately deep to the galea in order to maintain the maximal thickness and vascularity of the pericranium medial to the superior temporal lines bilaterally. Separation of these layers should only be taken to a level 1 cm above the supraorbital bar to preserve the rich vascular supply to the pericranium. Both superficial and deep layers of the temporalis fascia are incised. Dissection deep to the deep fascial layer minimizes the risk of injury to the frontalis branches of the facial nerve. Lateral subperiosteal dissection exposes the zygomatic process of the frontal bones, frontozygomatic sutures, and the upper lateral orbital rims. Medially, the subperiosteal dissection exposes the glabella, the frontonasal suture, and the superior orbital rims (Fig. 19).
Bone Dissection and Osteotomies Following a bifrontal craniotomy, the dura is dissected from the posterior wall of the frontal sinuses. The frontal sinuses are then cranialized and demucosalized with the high-speed drill and rongeurs. Careful dissection of the dura from the orbital surface of the frontal bones allows access for the orbital osteotomies. The orbital periosteum is dissected down from the roof and walls of the orbit. It has not been necessary in our experience to free the medial canthal ligaments. With protection of the orbital contents and dura, osteotomies are placed in the lateral wall, roof, and upper medial walls of the orbits and across the nasal bones to just in front of the crista galli. This bifronto–orbital osteotomy piece is freed with an osteotome and removed (Fig. 20).
Figure 17 Sagittal, coronal, and axial postcontrast T1-weighted MRI imaging of a patient with a low-grade, metastatic leiomyosarcoma to the sphenoid sinus and upper clivus presenting with progressive visual loss. This lesion was felt to be best approached via a transbasal route. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
286
DeMonte et al.
Figure 18 Artist’s representation of the improvement in the angle of vision following removal of the supraorbital osteotomy. This added 1.5 cm allows for a much greater angle of dissection, extending from the tip of the foramen magnum to the floor of the sella. There remains an area that is not well seen posterior and superior to the sella turcica. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
If an attempt at olfactory sparing is to be made, the dura is dissected along the side of the cribriform plate bilaterally to the level of the posterior ethmoidal artery. Under the microscope, a reciprocating saw is used to cut across the anterior planum sphenoidale at the level of the frontosphenoid suture. A generous cuff of nasal mucosa is preserved with the cribriform plate (12,29). Olfactory sparing is contraindicated for nasopharyngeal malignancies due to the neurotropism that many of these neoplasms exhibit. If, as is most often the case, olfactory preservation is not an issue the crista galli is removed and the dural sleeves of the olfactory groove are cut on either side and closed primarily. The planum sphenoidale is then opened and a wide sphenoidotomy is performed with a high-speed drill. The medial wall of the orbit is removed anteromedially to the optic nerves. One or both of the optic nerves may require decompression (Fig. 21). All but the lateral wall of the optic canal can be removed from this approach. The walls of the sphenoid are progressively removed from anterior to posterior. The sellar floor is removed. Access to the dorsum sella is blocked by the pituitary gland, and tumor in this area is not well seen although some degree of extradural elevation of the gland is possible. The cavernous sinuses and intracavernous ICA are exposed by the complete removal of bone on their medial aspects (Fig. 22). Although covered by a thin periosteal layer, bleeding from the cavernous plexus may occur and is
Figure 19 Artist’s representation of the skin incision and bifrontal craniotomy and biorbital osteotomy needed for the extended transbasal approach. The incision extends from the root of the zygoma bilaterally in a bicoronal fashion behind the hairline. The bifrontal craniotomy is created by placing entry holes at the keyhole region bilaterally and at the midline. Orbital osteotomies are raised by dividing the lateral orbital rims at the frontal zygomatic suture, the orbital roof, with protection of the dura and of the periorbital membrane and with divisions in the medial orbits under direct vision superiorly and medially. The final osteotomy is just in front of the crista galli and this frees the bilateral osteotomy. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
controlled by tamponade. Similarly, bleeding from the clival basilar venous plexus can be considerable but controlled with surgical packs. The clival tumor and bone are progressively removed in a superior to inferior direction with the foramen magnum being the lower limit of resection. At the level of the foramen magnum, the resection can be carried laterally up to the hypoglossal canals. Clival bone removal continues to include a margin of normal appearing bone if possible.
Tumor Resection The tumor is removed in an intralesional fashion. Depending on the consistency and vascularity, the tumor removal may be accomplished with microdissection and suction, ultrasonic aspiration, or with a high-speed diamond drill (Fig. 23).
Reconstruction and Closure Reconstruction begins by closure of any dural defects. Temporalis fascia or fascia lata are used as dural patches as necessary if primary closure is not possible. Usually, it is difficult if
Chapter 18: Clival Tumors
287
Figure 21 Intraoperative photograph of the transbasal approach in a patient with an extensive juvenile nasopharyngeal angiofibroma of the sphenoid sinus and clivus. The two orbits are anteriorly and laterally placed and all bone has been removed including that overlying the optic nerves bilaterally. The tumor is immediately anterior and inferior to the optic nerves. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission. Figure 20 Artist’s representation of the bifrontal craniotomy and biorbital osteotomy necessary for the extended transbasal approach. At this stage, the cribriform plate and ethmoid sinuses are removed and a wide sphenoidotomy performed to access the clivus. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
not impossible to get a watertight closure of the clival dura. In this case, as described above, a layer of fascia is placed intradurally with its edges extending beyond the dural opening and tacked down with sutures or hemoclips. A second extradural fascial layer is placed and reinforced with fibrin glue and autologous fat. The open ethmoid sinuses are packed with fat and then the pericranial flap is turned down to cover and reinforce the exposed dura and separate it from the paranasal sinuses. The bifronto–orbital osteotomy is replaced in its anatomic position and rigidly fixated with titanium plates. Care is taken not to compress the pericranial graft as it passes over the orbital soft tissues and nasal bones. The remainder of the closure follows standard techniques. Graduated, intermittent lumbar spinal drainage may be employed if the dural closure is tenuous.
Of the posterior transpetrous approaches, the petrosal or subtemporal–retrolabyrinthine approach is best suited for large posterolateral intradural tumor extensions [Fig. 24(A)– 24(C)] (39), while the ITF approach can be used for both extraand intradural tumor extensions (8). In the later approach, the structures of the jugular foramen can be identified and skeletonized with the high-speed drill. The extradural removal of tumors from around the exiting lower cranial nerves can result in neurologic improvement. The approach can be extended inferiorly by the posterior mobilization of the vertebral artery from the foramina transversaria of C1 and C2. In this way, the region of the odontoid process and arch of C1 can be exposed (40–43). Craniospinal fixation may be necessary. Total petrosectomy allows access from the sphenoid sinus to the foramen magnum and from the upper cervical spine to the intradural structures of the middle and posterior fossae. Destruction of the inner ear and complete mobilization of the facial nerve with their attendant morbidities are consequences of this approach. The problem of limited midline access remains (35).
Transpetrous Approaches In the presence of clival tumors with lateral extensions, adequate access for tumor resection may necessitate the use of a variety of transpetrous approaches. The amount of petrous bone resection is tailored to the location of the tumor. Anterior, posterior or rarely, total petrosectomy may be necessary (35). An anterior petrosectomy will access tumor extensions in and around the petroclival synchondrosis, the posterior cavernous sinus, and the prepontine and cerebellopontine cisterns. The limits of bone removal include the ICA laterally, the superior petrosal sinus medially, the inferior petrosal sinus inferiorly, and the internal auditory canal posteriorly (36). Sacrifice of the mandibular nerve and anterolateral displacement of the ICA allows for further inferior exposure of the clivus (37,38). Although a few appropriately located small tumors can be resected through this surgical corridor, an anteriorly based approach is typically required as a second stage.
Transcondylar Approach Tumors of the lower clivus with lateral extension to the atlanto–occipital joint, jugular foramen, or upper cervical spine are well addressed via this approach (44–50). Important advantages of the transcondylar approach include a short and wide surgical field, vascular control of the vertebral artery and an uncontaminated surgical field. The direct lateral exposure of the craniocervical junction afforded by this approach allows for a direct line of sight to the structures ventral to the cervicomedullary junction and upper cervical cord (51). This is an excellent approach for large lower clival chordomas. Neural decompression can be directly confirmed and craniocervical fixation performed should it be necessary. This typically saves the patient from immediate danger and allows time for a staged anterior approach, which in most cases is required for maximal tumor removal. The region of the occipital condyle may be accessed through either a posterolateral or an anterolateral approach (40,43). The
288
DeMonte et al.
Figure 22 Artist’s representation of the upper clivectomy that can be achieved via the transbasal approach. Both optic canals and medial cavernous sinus walls can be opened as can the anterior wall and floor of the sella turcica. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
posterolateral approach affords excellent anterior intradural midline access but does not allow a surgical corridor to the anterior extradural region. It is an excellent approach for clival tumors that penetrate the dura and extend posteriorly to compress the cervicomedullary region [Fig. 25(A)–25(D)]. It is typically used as a staged procedure, with a secondary anterior approach. The posterior midline access afforded allows for immediate craniocervical stabilization if necessary (Fig. 26). The posterolateral approach, sometimes called the far or extreme lateral approach is described elsewhere in this book. The anterolateral approach will be described below.
Positioning and Incision The patient is positioned in the supine position with the head slightly elevated and turned to the opposite side or in the direct lateral position. There are several options for the incision. We have used a large retroauricular C-shaped incision although an incision that begins along the upper anterior bor-
der of the sternomastoid muscle, curves across the mastoid process and then travels along the occipital crest to the external occipital protuberance can also be used. The exact choice of skin incision will be dependent on the extensions of the tumor, the need to combine this with another approach for resection or for concurrent occipitocervical fusion.
Soft Tissue Dissection Once the flap is elevated, the sternomastoid muscle is dissected along its anterior border to identify the jugular vein. The dissection continues superiorly and the sternomastoid muscle is detached from the mastoid tip and reflected posterolaterally (Fig. 27). The accessory nerve is identified in relation to the jugular vein and dissected superiorly as far as possible towards the skull base. Its distal portion, just prior to its penetration of the sternomastoid is also freed in order to mobilize the nerve inferiorly. The fatty tissue around the mid portion of the nerve is left undisturbed as a
Figure 23 Sagittal, coronal, and axial postcontrast MRI imaging following resection via a transbasal approach of the metastatic, low-grade leiomyosarcoma illustrated in Figure 17. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
Chapter 18: Clival Tumors
289
Figure 24 (A) Axial T2-weighted MRI reveals a large, intradural, laterally extending chordoma. The trigeminal nerve is displaced laterally. (B) Intraoperative photograph following completion of the petrous exposure. The petrous vein, trigeminal nerve, and VII–VIII nerve bundle are well identified as are the anterior inferior cerebellar artery and the tumor between the VII–VIII nerve complex and the trigeminal nerve. (C) Axial T2-weighted image following resection via the petrosal approach of this large intradural extension of chordoma. A secondary transmaxillary approach completed the resection in this patient. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
protective cushion around the nerve. The transverse process of the atlas is palpated just below and anterior to the tip of the mastoid process. The rectus minor, superior, and inferior oblique and the levator scapulae muscles are detached from the transverse process with attention to the vertebral artery,
which lays immediately deep to these muscles. As described by George, the vertebral artery is first exposed inferiorly between the transverse processes of C1 and C2. Here it is crossed anteriorly by the anterior branch of the C2 nerve root. The vertebral artery is then exposed above C1 and dissection
Figure 25 (A) Axial postcontrast MRI revealing a large intra- and extradural chordoma. (B) Intraoperative photograph of the posterolateral exposure of the condyle. The tumor is well identified as is the overlying hypoplastic vertebral artery. (C) Following resection and direct and complete decompression of the cervicomedullary junction. (D) Confirms neural decompression and removal of the posterolateral component of the tumor. A secondary transmandibular approach accessed and resected the remaining anterior disease. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
290
DeMonte et al.
Figure 28 This intraoperative photograph was taken following resection of the tumor identified in Figure 27. The vertebral artery is entirely skeletonized and the capsule of the atlantoaxial joint has been removed completely to identify the occipital condyle. A mastoidectomy and retrosigmoid craniectomy have been performed in order to achieve total tumor removal in this case of radiation-associated, osteogenic sarcoma. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
Figure 26 Lateral radiograph of the spine following occipital–cervical fusion in a patient who underwent complete condylar resection for removal of a clival chordoma. Used with permission.
of the C1–C2 joint. This fully exposes the lateral foramen magnum.
Bony Dissection and Osteotomies is carried posteriorly along the arch of C1 in a subperiosteal plane. The foramen transversarium of C1 is opened and the vertebral artery dissected free and translocated posteriorly. The C1 transverse process is resected and then the anterior arch of C1 is exposed and cleared off subperiosteally. Finally, the vertebral artery in its periosteal sheath is separated from the atlanto–occipital membrane and capsule
A standard mastoidectomy is performed and bone should be removed to unroof the lower half of the sigmoid sinus and jugular bulb. The superior extent of exposure is the jugular foramen. Tumor extensions above this level will require the addition of a transpetrous approach. A retromastoid craniotomy may be added if needed (Fig. 28). The extent of condylar resection depends on the extent of the tumor. Care must be taken in the identification of the hypoglossal nerve, which passes through the mid-portion of the condyle. Resection of greater than 50% of the condyle typically necessitates occipitocervical fixation. This can be performed at the same operation or at a secondary procedure with the patient maintained in a rigid collar in the interim.
Tumor Resection The tumor is resected using standard intralesional ultrasonic aspiration or suction and bipolar cautery. When significant encasement of the lower cranial nerves and the vertebral artery exists, cranial nerve monitoring serves as a useful adjunct. The dissection plane should proceed from normal to abnormal anatomy, especially in cases where prior surgery and treatment have created distorted anatomy and scar tissue.
Reconstruction and Closure Standard techniques for dural closure including primary dural closure and/or patch duraplasty are used to facilitate watertight dural closure in order to prevent CSF leak. Figure 27 Intraoperative photograph of a patient operated on via the anterolateral approach. The sternocleidomastoid muscle is reflected posteriorly. The accessory nerve is protected and the jugular vein is controlled in the vascular loop. The tumor is identified just below the tip of the mastoid overlaying the region of the condyle and vertebral artery. Source: From The Department of Neurosurgery, M.D. Anderson Cancer Center. Used with permission.
POSTOPERATIVE CARE AND FOLLOW-UP The most important aspect of the patients’ postoperative care is the identification and aggressive management of lower cranial neuropathy. All patients are evaluated by a speech
Chapter 18: Clival Tumors
pathologist in the immediate postoperative period and a swallowing evaluation performed. A modified barium swallow or direct laryngoscopy may be recommended. Only when deemed safe, oral intake is allowed. Disturbances of ocular movement are managed initially by patching and subsequently by prisms and extraocular muscle surgery under the supervision of an ophthalmologist. A noncontrast CT of the brain is done on the first postoperative day to identify any subclinical complications. A baseline postoperative MRI is done within 48 hours after surgery. If clinically indicated, dynamic imaging of the craniocervical junction with flexion– extension radiographs is obtained. Close follow-up with clinical evaluation and imaging is essential, the frequency of which is dependent on tumor pathology. For patients with chordoma or chondrosarcoma, if no tumor is visualized on baseline postoperative imaging, additional imaging is obtained at 3-month intervals for the first year and then every 6 months for the next 2 years and yearly thereafter. When residual disease is identified, the patient should be referred to a specialized radiation treatment facility, ideally one that has proton beam capability as well as radiation oncologists familiar with the treatment of disease processes that affect the clivus. Referral to medical oncology may also be necessary. Management algorithms for residual or recurrent local disease is complex and beyond the scope of this chapter. Multiple variables should be analyzed including the clinical condition of the patient; the size and location of the recurrence; the tumor histology at the initial operation; the extent of the previous resection; and the dose, field, and fractionation scheme of prior radiation therapy. REFERENCES 1. Alfieri A, Jho HD. Endoscopic endonasal cavernous sinus surgery: An anatomic study. Neurosurgery. 2001;48:827–836. 2. Cavallo LM, Messina A, Cappabianca P, et al. Endoscopic endonasal surgery of the midline skull base: Anatomical study and clinical considerations. Neurosurg Focus. 2005;19:E2. 3. Cavallo LM, Messina A, Gardner P, et al. Extended endoscopic endonasal approach to the pterygopalatine fossa: Anatomical study and clinical considerations. Neurosurg Focus. 2005;19:E5. 4. De Oliveira E, Rhoton AL, Peace DA. Microsurgical anatomy of the region of the foramen magnum. Surg Neurol. 1985;24:293– 352. 5. Oot RF, Melville GE, New PF, et al. The role of MR and CT in evaluating clival chordomas and chondrosarcomas. Am J Roetgeol. 1988;151:567–575. 6. Meyers SP, Hirsch WL Jr, Curtin HD, et al. Chordomas of the skull base: MR features. Am J Neuroradiol. 1992;13:1627–1636. 7. Maira G, Pallini R, Anile C, et al. Surgical treatment of clival chordomas: The transsphenoidal approach revisited. J Neurosurg. 1996;85:784–792. 8. Laws ER Jr. Transsphenoidal surgery for tumors of the clivus. Otolaryngol Head Neck Surg. 1984;92:100–101. 9. Lalwani AK, Kaplan MJ, Guti PH. The transsphenoethmoid approach to the sphenoid sinus and clivus. Neurosurgery. 1991;31:1008–1014. 10. Crockard HA, Sen CN. The transoral approach for the management of intradural lesions at the craniovertebral junction: Review of 7 cases. Neurosurgery. 1991;28;88–98. 11. Menezes AH, VanGilder JC. Transoral-transpharyngeal approach to the anterior craniocervical junction. Ten-year experience with 72 patients. J Neurosurg. 1988;69:895–903. 12. Beals SP, Jogannic EF, Hamilton MG, et al. Posterior skull base transfacial approaches. Clin Plast Surg. 1995;22:491–511. 13. Catalano PJ, Biller HF. Extended osteoplastic maxillotomy. A versatile new procedure for wide access to the central skull base and infratemporal fossa. Arch Otolaryngol Head Neck Surg. 1993;119:394–400.
291
14. James D, Crockard HA. Surgical access of the base of the skull and upper cervical spine by extended maxillotomy. Neurosurgery. 1991;29;411–416. 15. Wei WI, Lam KH, Sham JS. New approach to the nasopharynx: The maxillary swing approach. Head Neck. 1991;13:200–207. 16. Couldwell WT, Sabit I, Weiss MH, et al. Transmaxillary approach to the anterior cavernous sinus: A microanatomic study. Neurosurgery. 1997;40:1307–1311. 17. Rabadan A, Conesa H. Transmaxillary-transnasal approach to anterior clivus: A microsurgical model. Neurosurgery. 1992:30(4):473–481. 18. Sabit I, Schaefer SD, Couldwell WT. Extradural extranasal combined transmaxillary transsphenoidal approach to the cavernous sinus: A minimally invasive microsurgical model. Laryngoscope. 2000;110;286–291. 19. Uttley D, Moore A, Archer DJ. Surgical management of midline skull base tumors: A new approach. J Neurosurg. 1989;71:705– 710. 20. DeMonte F, Hanna EY. Transmaxillary exploration of the intracranial portion of the maxillary nerve in malignant perineural disease. J Neurosurg. 2007;107:672–677. 21. Sabit I, Schaefer SD, Couldwell WT. Modified infratemporal fossa approach via lateral transantral maxillotomy: A microsurgical model. Surg Neurol. 2002;58:21–31. 22. Biller HF, Shugar JM, Krespi YP. A new technique for wide-field exposure of the base of the skull. Arch Otolaryngol. 1981;107:698– 702. 23. Krespi YP, Har-El G. Surgery of the clivus and anterior cervical spine. Arch Otolaryngol Head Neck Surg. 1988;114:73–78. 24. Krespi YP, Sisson GA. Skull base surgery in composite resection. Arch Otolaryngol. 1982;108:681–684. 25. Krespi YP, Sisson GA. Transmandibular exposure of the skull base. Am J Surg. 1984;148:534–583. 26. Ammirati M, Bernardo A. Analytical evaluation of complex anterior approaches to the cranial base: An anatomical study. Neurosurgery. 1998;43:1398–1407. 27. Ammirati M, Ma J, Cheatham ML, et al. The mandibular swingtranscervical approach to the skull base: Anatomical study. [technical note] J Neurosurg. 1993;78:673–681. 28. DeMonte F, Diaz E, Callender D, et al. The transmandibular, circumglossal, retropharyngeal approach for chordomas of the clivus and upper cervical spine. Neurosurg Focus. 2001;10:1–5. 29. Feiz-Erfan I, Spetzler RF, Horn EM, et al. Proposed classification for the transbasal approach and its modifications. Skull BaseInterdisciplin Approach. 2008;18:29–48. 30. Fujitsu K, Saijoh M, Aoki F, et al. Telecanthal approach for meningiomas in the ethmoid and sphenoid sinuses. Neurosurgery. 1991;28:714–719. 31. Gay E, Sekhar LN, Rubinstein E, et al. Chordomas and chondrosarcomas of the cranial base: Results and follow-up of 60 patients. Neurosurgery. 1995;36:887–896. 32. Kawakami K, Yamanouchi Y, Kawamura Y, et al. Operative approach to the frontal skull base: Extensive transbasal approach. Neurosurgery. 1991;28:720–724. 33. Sekhar LN, Nanda A, Sen CN, et al. The extended frontal approach to tumors of the anterior, middle and posterior skull base. J Neurosurg. 1992;76:198–206. 34. Terasaka S, Day JD, Fukushima T. Extended transbasal approach: Anatomy, technique, and indications. Skull Base Surg. 1999;9:177–184. 35. Leonetti JP, Reichman OH, Al-Mefty O, et al. Neurotologic considerations in the treatment of advanced clival tumors. Otolaryngol Head Neck Surg. 1992;107:49–56. 36. Kawase T, Shiobara R, Toya S. Anterior transpetrosal transtentorial approach for sphenopetroclival meningiomas: Surgical method and results in 10 patients. Neurosurgery. 1991;28:869– 876. 37. Harsh GR IV, Sekhar LN. The subtemporal, transcavernous, anterior transpetrosal approach to the upper brain stem and clivus. J Neurosurg. 1992;77:709–717. 38. Sen CN, Sekhar LN. The subtemporal and preauricular infratemporal approach to intradural structures ventral to the brain stem. J Neurosurg. 1990;73:345–354.
292
DeMonte et al.
39. Blevins NH, Jackler RK, Kaplan MJ, et al. Combined transpetrosal-subtemporal craniotomy for clival tumors with extension into the posterior fossa. Laryngoscope. 1995;105:975– 982. 40. Carpentier A, Blanquet A, George B. Suboccipital and cervical chordomas: Radical resection with vertebral artery control. Neurosurg Focus. 2001;10(3):1–6. 41. George B, Dematons C, Cophignon J. Lateral approach to the anterior portion of the foramen magnum. Application to surgical removal of 14 benign tumors: Technical note. Surg Neurol. 1988;29:484–490. 42. George B, Laurian C. Surgical approach to the whole length of the vertebral artery with special reference to the third portion. Acta Neurochir. 1980;51:259–272. 43. George B, Lot G. Anterolateral and posterolateral approaches to the foramen magnum: Technical description and experience from 97 cases. Skull Base Surg. 1995;5:9–19. 44. Babu RP, Sekhar LN, Wright DC. Extreme lateral transcondylar approach: Technical improvements and lessons learned. J Neurosurg. 1994;81:49–59.
45. Borba LA, Al-Mefty O. Skull-base chordomas. Contemp Neurosurg. 1998;20:1–6. 46. Canalis RF, Martin N, Black K, et al. Lateral approach to tumors of the craniovertebral junction. Laryngoscope. 1993;103:343–349. 47. George B, Archilli M, Cornelius JF. Bone tumors at the craniocervical junction. Surgical management and results from a series of 41 cases. Acta Neurochir (Wien). 2006;148:741–749. 48. Lang DA, Neil-Dwyer G, Iannotti F. The suboccipital transcondylar approach to the clivus and cranio-cervical junction for ventrally placed pathology at and above the foramen magnum. Acta Neurochir (Wien). 1993;125:132–137. 49. Sen CN, Sekhar LN. An extreme lateral approach to intradural lesions of the cervical spine and foramen magnum. Neurosurgery. 1990;27:197–204. 50. Sen CN, Sekhar LN. Surgical management of anteriorly placed lesions at the craniocervical junction—An alternative approach. Acta Neurochir. 1991;108:70–77. 51. Wen H, Rhoton AL Jr, Katsuta T, et al. Microsurgical anatomy of the transcondylar, supracondylar, and paracondylar extensions of the far-lateral approach. J Neurosurg. 1997;87:555–585.
19 Tumors of the Anterior Skull Base Vijayakumar Javalkar, Bharat Guthikonda, Prasad Vannemreddy, and Anil Nanda
SURGICAL ANATOMY
noidale, anterior clinoid process, diaphragma sellae, orbital roof, and tuberculum sellae. The most common sites are the olfactory groove or tuberculum sellae. Any of these locations can give rise to meningiomas that invade the underlying bone and paranasal sinuses. Malignant meningiomas comprise approximately 1% to 7% of all meningiomas with another 5% being atypical (3). There are nine subtypes of meningioma with low risk of recurrence (WHO grade I). Rhabdoid, papillary, and anaplastic (malignant) are classified as WHO grade III tumors and signify a higher likelihood of aggressive behavior and recurrence. Atypical, clear cell, and chordoid are intermediate grade tumors (WHO grade II). Unlike meningiomas with low risk of recurrence, malignant meningiomas are more common in males.
The floor of the anterior cranial fossa is formed by the orbital plates of the frontal bone, the cribriform plate of the ethmoid bone, and the lesser wings and anterior body of the sphenoid. Posteriorly, it is bounded by the lesser wings of the sphenoid and anterior margin of the chiasmatic sulcus. It typically contains irregular crevices upon which the frontal lobes rest. The medial portion of the anterior cranial fossa is the barrier between the intracranial space superiorly and the nasal cavity inferiorly. The medial anterior cranial fossa (Fig. 1) is bordered by the paranasal sinuses: anteriorly by the frontal sinus, laterally by the ethmoid sinuses, and posteriorly by the sphenoid sinus (1). The central part of the floor of the anterior cranial fossa is formed by the ethmoid bone. Tumors arising from the skull base may erode this part of the anterior skull base resulting in intracranial or sinus extension. The cribriform plate is the perforated central portion of the ethmoid. It contains 30 to 50 foramina (ranging 0.1–0.5 mm in diameter), which transmit the olfactory nerves (1). The crista galli is a sharp projection of the ethmoid bone in the midline, which gives attachment to the falx cerebri. The olfactory grooves on either side of the crista galli support the olfactory bulbs. Lateral to the olfactory grooves, the floor continues as the roof of the ethmoidal sinuses (fovea ethmoidalis). The foramen caecum is a small aperture situated between the crista galli and frontal crest, through which a small vein from the nasal cavity traverses to the superior sagittal sinus. The dura mater over the cribriform plate is thin, and care should be taken when reflecting the dura from this part of the floor. Lateral to either olfactory groove are the internal openings of the anterior and posterior ethmoidal foramina. The anterior ethmoidal foramina transmit the anterior ethmoidal vessels and the nasociliary nerve. The posterior ethmoidal foramen transmits the posterior ethmoidal vessels and nerve.
Esthesioneuroblastoma Esthesioneuroblastoma was first described by Berger et al. in 1924 (4). Schall and Lineback described the first cases in America in 1951 (5). They arise from the olfactory neuroepithelium, and the incidence is 3% of all intranasal tumors (6). Esthesioneuroblastoma can invade the cribriform plate, paranasal sinuses, intracranial cavity, and orbits. The most common site of metastasis is cervical lymph nodes, while sites of distant metastasis such as lungs, brain and bone are less common (7). It is most common in the fifth decade of life (8). A bimodal distribution has been shown at peaks in age groups 11 to 20 and 51 to 60 years (9,10). The sex distribution is almost equal. The Kadish staging system is the most commonly used system (Table 1) (11). Surgery followed by radiation has been widely adopted as the standard of care in Kadish stages A and B disease. For patients with Kadish C disease, the use of combined modality treatment with surgery followed by radiation and chemotherapy has been adopted (12). In patients with advanced and high-grade esthesioneuroblastoma, the option of platinum-based chemotherapy should be considered (13).
Chondrosarcoma
REGIONAL PATHOLOGY AND DIFFERENTIAL DIAGNOSIS
Chondrosarcoma constitutes about 6% of all the skull base tumors (14). They are thought to arise from the persistent islands of embryonal cartilage that occur near the cranial base synchondroses (15). The majority of them arise from the middle cranial fossa (64%) followed by the anterior cranial fossa (14%) and 7% originate from the posterior fossa (16). They should be distinguished from chordomas, especially the chondroid variety of chordoma, which may mimic a chondrosarcoma (17). Staining for epithelial membrane antigen and cytokeratin (chondrosarcomas do not stain positive) should be carried out to distinguish them from chordoma. Four primary histologic types have been reported: conventional (hyalin and myxoid), dedifferentiated, clear cell, and mesenchymal (18,19). Most cranial base chondrosarcomas
The anatomic structures that are intimately related to the anterior cranial fossa are the frontal bone; frontal, ethmoid, and sphenoid sinuses; nasal cavity; nasopharynx; and orbit and its contents. Any tumor, either benign or malignant, arising from these anatomic structures can erode through the anterior cranial fossa floor. The following is a summary of pathologies that often affect the anterior skull base.
Meningioma Meningiomas often arise from the dura mater of the anterior cranial fossa. Anterior cranial fossa meningioma constitutes about 40% of all the intracranial meningiomas (2). They originate from the dura of the olfactory groove, planum sphe293
294
Javalkar et al.
cases and 26% recurrence-free survival for reoperated cases at 10 year follow-up (22). Various forms of radiation treatment like high-dose protons or charged particles, such as carbon ions, helium or neon and radiosurgery using the gamma-knife or cyber knife is used to complement surgery. Chemotherapeutic trials with agents such as imatinib have also been tried and the results of a phase II study on imatinib are awaited (30).
Nasopharyngeal Carcinoma
Figure 1 Anterior cranial fossa floor. Neural–dural transition at the medial anterior cranial base. Source: From Ref. 1.
are of the low-grade variety and grow slowly. Dedifferentiated and mesenchymal chondrosarcomas are much more aggressive, but fortunately are rare (20,21). Magnetic resonance imaging (MRI) is the best study for preoperative diagnosis. On short repetition time (TR)/echo time (TE) MRI scans, chondrosarcomas generally exhibit low signal intensity. However, on long TR/TE MRI scans, they generally have very high signal intensity. After the administration of gadolinium, they show a marked enhancement in either a heterogeneous or homogeneous pattern. In one study, with particularly long follow-up, a recurrence free survival of 32% at 10 years was reported (22).
Chordomas Chordomas are slow-growing neoplasms arising from the remnants of primitive notochord. They are predominantly located in the sacrococcygeal region and the clivus. Fifty percent of chordomas involve the sacrococcygeal region, 35% occur at skull base near the spheno–occipital area, and 15% are found in the vertebral column (23). Chordomas typically present in adult males. Chordomas in children and adolescents are very uncommon, and account for ≤5% of all chordomas (24). Craniocervical chordomas most commonly involve the dorsum sellae, clivus, and nasopharynx. Chordomas are divided into conventional, chondroid, and dedifferentiated types. Conventional chordomas are the most common. Chondroid chordomas contain both chordomatous and chondromatous features and this histology is associated with better prognosis (25,26). Dedifferentiation or sarcomatous transformation occurs in 2% to 8% of chordomas and this can develop at the onset of the disease or later (27,28). Surgical treatment plays a major role in the management of chordoma. The main prognostic factor for local failure, and possibly survival, is quality of surgical margins (29). Tzortzidis et al. reported a 42% recurrence-free survival for primary Table 1 Kadish Staging—Esthesioneuroblastoma Stage A Stage B Stage C
Tumor confined to the nasal cavity Tumor confined to the nasal cavity and paranasal sinuses Tumor extends beyond the nasal cavity and paranasal sinuses, including involvement of cribriform plate, base of skull, orbit, or intracranial cavity
Nasopharyngeal carcinoma is a nonlymphomatous squamous cell carcinoma, which is uncommon in most countries and its age-adjusted incidence for both sexes is less than one per 100,000 population (31). In endemic areas like southern China, the incidence is much higher at 15–50/100,000 per year (32). The incidence is high among Chinese people who have immigrated to Southeast Asia or North America, but is lower among Chinese people born in North America than those born in China (31). Epstein-Barr virus is implicated in the pathogenesis of nasopharyngeal carcinoma and even the premalignant lesions harbor EBV (33). According to WHO classification, these are classified into three types. Type I includes typical keratinizing squamous cell carcinoma, type II includes nonkeratinizing squamous cell carcinoma, and type III includes undifferentiated carcinoma (34). Definitive diagnosis is made by endoscopic guided biopsy (32). Disease staging is based on the International Union against Cancer and American Joint Committee on Cancer staging system (AJC/UICC). The extent of nasopharyngeal carcinoma as per the TNM staging is the most important prognostic factor (31). Radiation treatment is the mainstay of treatment. Stage I and IIA lesions are treated by RT alone, while stages III and IVA, B lesions are treated by RT with concurrent chemotherapy followed by adjuvant chemotherapy [I, A]. Concurrent chemotherapy could also be considered for stage IIB disease [III, B]. The standard agent used in concurrent chemotherapy is cisplatin [I, A]. Adjuvant chemotherapy on its own has not been documented to confer survival advantage, adjuvant cisplatin and 5-fluorouracil combined with concurrent cisplatin RT has been demonstrated to be beneficial [I, A] (34). Gene therapy and immunotherapy are also being studied (35,36).
Malignant Melanoma Mucosal malignant melanoma (MM) accounts for 1% of all melanomas (37). Of all the mucosal malignant melanomas, sinonasal melanoma is the most common site. The most common sinonasal malignant melanoma sites are the lateral wall and the inferior turbinate of nasal cavity, maxillary sinus, and the ethmoid sinus (38). It is seen in fifth to eighth decade and occurs equally in men and women (39). Prognosis is poor, and the 3-year overall survival for malignant melanoma was only 28.2% in one study (38).
Lymphoma Non-Hodgkin lymphomas of the sinonasal tract are uncommon malignancies in the United States, representing approximately 1.5% of all lymphomas (40). The incidence has been reported to be higher in Asians and South American countries (41). Diffuse large cell B-cell lymphoma is the most common variety (42). Skull base involvement associated with sinonasal lymphoma is rare (43). It is associated with Epstein–Barr virus, and is also known as lethal midline granuloma syndrome (44). Lymphoma usually presents as a midfacial destructive lesion with nasal obstruction, epistaxis, ulceration, fever, chills, and weight loss (39). Fine needle aspiration or
Chapter 19: Tumors of the Anterior Skull Base
biopsy can reveal the diagnosis. If fine needle aspiration or biopsy is unavailable or if the fine needle aspiration is insufficient for diagnosis, then nasal biopsy is suggested (42). T-cell lineage extranodal NHL is associated with poor prognosis (45). Traditionally, radiation was considered the main stay of treatment for low-grade histology and stage I and stage II disease. In recent studies, combined treatment with chemotherapy and radiation has shown to improve survival (40,46–48).
Angiofibroma Angiofibromas are benign highly vascular tumors that occur in adolescent males. The classical presentation is nasal obstruction and epistaxis. The common routes of invasion are anterior infiltration of nasal cavity, anterolateral erosion of the posterior maxillary sinus, and anterosuperior destruction of the ethmoid air cells (49). They may also extend laterally into the pterygopalatine fossa (42). Imaging studies may reveal classical anterior bowing of the posterior wall of maxillary sinus (Holman–Miller sign). MRI better delineates the extent of tumor and spatial relationship with surrounding vital structures. Surgical resection after preoperative embolization is the main stay of treatment. Combined approaches are needed for those with intracranial extension. Radiation treatment has also been used as a treatment modality for angiofibromas (50).
Adenoid Cystic Carcinoma Adenoid cystic carcinoma is a malignant tumor that occurs most frequently in the parotid gland. They account for 10% to 15% of paranasal sinus malignancies (42). They have a high propensity for perineural spread. These tumors grow silently and present at a late stage due to submucosal extension. Adenoid cystic carcinoma usually arises in the sixth decade of life with nearly equal occurrence in males and females (51). Hematogenous metastasis is known to occur to the lungs after clinical progression of several years (52,53). Histologically, three growth patterns are recognized: cribriform, tubular, and solid. Solid growth pattern is associated with worse prognosis (54). Indeed, the worst prognostic factor in patients with adenoid cystic carcinoma is involvement of the skull base (55). Aggressive treatment with combined surgery and irradiation therapy is often inadequate to achieve local control (56). High-dose, conformal proton beam radiation therapy has been tried in patients with adenoid cystic carcinoma of the skull base with some success (55). Other modalities like Gamma Knife radiosurgery has been tried in the management of recurrent salivary gland tumors involving the skull base (57,58).
CLINICAL ASSESSMENT Clinical assessment should include detailed clinical history and physical examination including a comprehensive head and neck examination. The presence or absence of palpable lymph nodes in the neck should be determined. If the lymph nodes are palpable, then the location, size, mobility, and relationship to the carotid and other adjacent structures should be noted. Anterior skull base tumors have a propensity to involve the olfactory nerves, the visual system, including the orbit, optic nerves, and optic chiasm. Commonly overlooked, anosmia can be associated with significant lifestyle modifications for the patient and the family (59). A formal visual acuity assessment, funduscopic evaluation, and visual field testing are essential upon diagnosis. Surgery or radiation treatments
295
can both lead to worsening of visual function and thus pretreatment assessment is crucial. Preoperative full extraocular motility correlates well with the likelihood of preserving orbital function (42). Evaluation by a multidisciplinary team consisting of neurosurgeon, head and neck surgeon, medical oncologist, and radiation oncologist is essential for delivery of optimal care for anterior skull base malignancies.
DIAGNOSTIC IMAGING MRI High quality MRI scan is the examination of choice. The advantages with MRI are superior soft tissue resolution, multiplanar imaging, and no ionizing radiation. It not only shows the extent of tumor, but also the relationship with important neurovascular structures like cranial nerves and blood vessels. Thin section, coronal, fast, spin echo T2-weighted images are useful for assessing the presence or absence of subtle anterior skull base penetration, whereas coronal postgadolinium T1-weighted images with fat saturation most sensitively assess dural, leptomeningeal, and parenchymal extension (42). MRI allows evaluation of the potential perineural spread of tumor along the branches of the trigeminal nerve (60,61). Even though nerve enhancement and thickening is not specific to perineural spread, such an observation is more likely in the setting of head and neck cancer due to perineural spread rather than inflammation. MRI has a higher sensitivity and specificity in detecting perineural spread than CT (62). MRI appearances may indicate the origin of anterior skull base tumors (63). MRI is especially useful in evaluating the orbit and its contents, but peritumoral edema can lead to abnormal signal intensity within ocular muscles and abnormal contrast enhancement of muscles, which may simulate neoplasm on MR imaging. This false-positive result is less commonly seen with a CT scan (64). Orbital infiltration is generally limited by the tough fibrous periorbita, which functions as an excellent barrier against the spread of neoplasm. Unfortunately, the periorbita is not readily identifiable even in high-resolution MR imaging or thin-section CT. MR imaging is also useful to study vascular encasement. In one study, the authors found that a single criterion of involvement of 270 degree or more of the circumference was accurate in predicting the surgeon’s ability to peel the tumor off the carotid artery (65).
CT A thin-section, coronal CT scan can be helpful in evaluating the cribriform plate for invasion (42). The following findings might suggest skull base invasion by tumors on CT scan: foraminal enlargement, bony thinning, erosion, and displacement (65). Perineural spread on a CT is demonstrated by the enlargement of the neural foramina and or soft-tissue infiltration of the fat adjacent to or within the neural foramina (66– 68). CT evidence of perineural spread is often an inference whereas there can be direct visualization of enhancement of the nerve with MR imaging (69). CT images can be fused with PET scans and the PET-CT fusion technique might be useful in evaluating skull base malignancies (70).
PET Scan Preoperative PET scan in anterior skull base tumors has a very low yield (71). PET-CT scan may show better results for the detection of lymph node metastasis than PET or CT alone (72,73). In long-term surveillance of patients, obtaining a preoperative PET scan may be helpful in that any increase in
296
Javalkar et al.
unexpected activity can be evaluated (42). The important limitations associated with PET scans are due to false-positive and false-negative results following treatment. During the first three months following radiotherapy, there is false-negative uptake of FDG (74). False-positive results may be due to early postoperative changes, inflammation at the primary site, or regional lymph nodes (71).
Vascular Encasement In cases where the major blood vessels are involved by the tumor, resection carries a risk of cerebrovascular insufficiency. The imaging studies that are useful in such cases are MR angiograms, conventional angiography, temporary balloon occlusion, and Xe-CT CBF studies (69). Anatomic imaging techniques like conventional angiogram and MRI are less reliable in predicting the risk of stroke. Preoperative carotid artery temporary balloon occlusion with Xe-CT CBF scanning provides the most reliable measure of the collateral carotid artery circulation (75). Even in cases where a successful preoperative assessment is completed, the risk of developing stroke is 3% to 8% (59). Conventional angiography is also used to study the overall vascularity of the tumor and occasionally endovascular embolization of the arterial tumoral feeders is helpful.
Biopsy Whenever possible, an endoscopic transnasal biopsy is obtained (59,75). This will help to differentiate certain tumors, like primary lymphomas, which have a nonconventional or highly specific treatment strategy (42).
SURGICAL TECHNIQUE The goals of anterior skull base surgery are resection of tumor with negative margins. Occasionally, en-bloc resection is sought (especially for primary bone lesions); however, this is rarely practical in anterior skull base tumors. Minimal brain retraction, preservation of neurologic function, and improved aesthetic outcome are essential tenets to anterior skull base surgery (42). Reconstruction of the anterior cranial base is also an extremely important part of the surgical strategy to reduce postoperative CSF leak and infections. The preoperative plan for closure is equally important to the plan for approaching and resecting the tumor. The concept of clear margins applies to anterior skull base malignancies similar to the other head and neck malignancies. If possible, negative margins should be obtained; however, given the critical structures in the area, this is often not feasible (76). The importance of histologically negative margins is universally accepted (38,77).
APPROACHES FOR ANTERIOR SKULL BASE MALIGNANCIES Craniofacial Resection Surgery remains the main modality of treatment for anterior skull base malignancies. The craniofacial approach constitutes the main surgical strategy. In 1954, Smith reported a case in which the craniofacial approach was used for the first time (78). It was only after Ketcham’s publication in 1966 that the craniofacial approach became the standard for the surgical treatment of malignancy of the anterior skull base reaching the lamina cribrosa (79). The important contraindications for craniofacial resection are significant invasion of the brain, invasion of the optic chiasm, invasion of both orbits,
Figure 2 Bicoronal skin incision extending from one zygomatic arch to the other. Source: From Ref. 89.
invasion of the internal carotid artery, cavernous sinus extension, and elderly patients with associated comorbid medical illness (42,80,81). The advances in imaging techniques, neuronavigation, operating microscope, sophisticated skull base approaches, and reconstructive techniques have all contributed to the overall improvement of results. Craniofacial resection is safe and effective treatment option for patients with malignant tumors of the skull base. In the International Collaborative Study of Craniofacial Surgery for malignant skull base tumors, the mortality rate was 4% and postoperative complications were noted in 33% of cases (77).
Preoperative Preparation Broad-spectrum antibiotics are administered at the time of induction of general anesthesia. The use of broad-spectrum antibiotics in craniofacial resections has been shown to reduce the incidence of infectious complications (82–86). There are no definite guidelines regarding the use of lumbar drains at the time of surgery. Most feel that drainage of CSF achieves brain relaxation and thus reduces the retraction of the brain (75,87). Others believe that routine lumbar drain is not recommended and sacrifice of the nerves and fenestration of the peri-olfactory cisterns enable adequate drainage of CSF (88).
Incision The standard bicoronal frontal incision extends from the tragus of one ear to the other behind the hairline (Fig. 2). The scalp incision is extended down to the plane between the epicranial aponeurosis and the pericranium. The posterior scalp flap is retracted several centimeters to expose extrapericranium for reconstruction so that sufficient length of flap can be harvested. The scalp edges are secured with Raney clips. The pedicled galeal-pericranial flap is considerably stronger than the pericranial flap and it is the local flap of choice when reconstructing a skull base defect (75).
Soft Tissue Dissection The anterior flap is elevated to expose the supraorbital ridge, glabella, and the upper half of nasal bones (Fig. 3). Care should be taken to preserve the supratrochlear and supraorbital vessels, which supply the flap. After elevation of the flap, the galeal-pericranial flap is dissected and wrapped in bacitracin-soaked gauze throughout the procedure. A 1 cm
Chapter 19: Tumors of the Anterior Skull Base
297
Figure 3 Dissection of pericranium in preparation for bifrontal craniotomy. Source: From Ref. 89.
cuff of galea should be preserved on the anterior flap for subsequent scalp closure. After reflection of the scalp flap, proposed bony cuts are marked on the anterior wall of frontal sinus and bone.
Bony Dissection and Osteotomy A single burr hole is made in the midline 5 cm above the glabella. Bone cuts are made with a drill bilaterally up to the midorbit without entering the frontal sinus. Frontal sinus cuts are made using 2 mm cutting burr through the anterior and posterior wall of the frontal sinus. This strategy reduces the incidence of dural tears adjacent to the frontal sinus (75). A small frontal craniotomy is largely favored. After the bony cuts, the frontal bone flap is elevated. The remainder of the posterior wall of the frontal sinus is removed using a rongeur. Care should be taken to remove mucosa of the sinus and cranialize the frontal sinus by excising the posterior wall. Any bleeding from the sagittal sinus can be controlled with tamponade and gelfoam. The technique of the craniotomy might vary. While elevating the dura from the base of anterior cranial fossa, care must be taken, especially while separating it from the crista galli. Once the dura is separated, the crista galli can be excised by a rongeur. The dura is further separated from the cribriform plate and planum sphenoidale. Undue retraction should be avoided. If a lumbar drain was instituted, draining 30 cc of CSF will relax the brain at this stage, allowing for easier extradural dissection. The bone cuts in the floor of the anterior cranial fossa are completed using a fine cutting burr (Fig. 4). The tumor can be approached from the inferior aspect by transfacial approaches. Modified Weber–Ferguson incision is favored because of the better cosmetic result (75). Facial incision may be avoided in cases of selected tumors of the superior nasal vault or ethmoid sinuses if adequate exposure and resection can be accomplished transcranially. After the appropriate transfacial approach, the tumor is removed, en bloc if possible, using both the transfacial and transcranial routes.
Subcranial Approach The concept of the subcranial approach to anterior skull base was first introduced by Raveh, in 1978, initially for the treatment of trauma to the anterior skull base, and later it was applied to skull base tumors (91,92). This approach combines a bifrontal craniotomy with naso-orbital osteotomies. It provides adequate exposure of the orbital, sphenoethmoidal, and clival regions as well as nasal and paranasal sinuses (93). The
Figure 4 Illustration of the bone cuts around the cribriform plate. Source: From Ref. 90.
advantages of this procedure are minimal frontal lobe retraction and no facial incisions. A potential disadvantage of the subcranial approach, when it is used in resection of malignant tumors, is osteomyelitis or osteoradionecrosis of the disconnected and replaced bones that includes the medial orbital rim, glabella, and part of the nasal bone (42).
Midfacial Degloving Approach Midfacial degloving (MFD) approach avoids facial incisions. It can be combined either with bifrontal craniotomy or subcranial approaches for tumor resection. It provides adequate exposure to nasal cavities, ethmoid complexes, maxillary sinuses, ethmoid roof, cribriform plate, sphenoid sinuses, and nasopharynx (94). Extended unilateral maxillotomy approach to the skull base can be carried out through MFD approach (95).
Orbitozygomatic Approach This approach was first described by Pellerin et al. and Hakuba et al. (96,97). Removal of the zygomatic arch, lateral orbit, and part of the zygomatic body with a standard pterional craniotomy can give access to the parapharyngeal space and the floor of the middle cranial fossa (42). The orbitozygomatic approach is also recommended for skull base tumors that involve the cavernous sinus (98).
SURGICAL APPROACHES FOR ANTERIOR SKULL BASE MENINGIOMAS The surgical approach depends on the location of the tumor. The most common surgical approaches are variants of either standard pterional or subfrontal (2,99). Extended frontal approaches to anterior skull base tumors have been proposed (100). This approach is a modification of the transbasal approach of Derome. The addition of bilateral orbitofrontal or orbitofrontoethmoidal osteotomy (Figs. 5–7) improves the exposure of midline lesions, while minimizing
298
Javalkar et al.
Figure 5 Extent of bone removal in fronto-orbital osteotomy. Source: From Ref. 89.
the need for frontal lobe retraction. Compared with a conventional bifrontal craniotomy, the addition of fronto-orbital osteotomy increases the surgical exposure while minimizing frontal lobe retraction. In one study, the angle of exposure increased markedly with an average gain of 76% for the anterior cranial fossa (89). Otani et al. proposed selective extradural anterior clinoidectomy for surgical management of tuberculum sellae meningiomas (101). They recommend this procedure for complete resection of tuberculum sellae meningiomas extending to the surrounding anatomic structures. Arai et al. proposed a transcranial–transsphenoidal approach for tuberculum sellae meningiomas (102). They stressed that this approach provides good access to a tumor extending inferomedially to the optic nerve. The other approaches that are published in the literature are frontal sinus approach to olfactory groove meningiomas, and orbital roof craniotomy
Figure 7 Basal view of skull showing the extent of bone removal in orbitofrontoethmoidal osteotomy. Source: From Ref. 89.
via an eyebrow incision (103,104). Endoscopic endonasal and endoscopic glabellar approaches to the anterior skull base has been described as well (105,106).
Endoscopic Approaches Endoscopic resection of anterior skull base tumors was described by Casiano to resect esthesioneuroblastoma (107). This technique has been used to resect other tumors like angiofibromas, ethmoid carcinomas, fibro-osseous lesions, and chondro-osseous lesions. Endoscopic approaches avoid facial incisions and bony cuts thus offering a cosmetic advantage. There are concerns regarding the oncologic safety of endoscopic tumor removal in case of malignancies. Long-term outcome results of endoscopic tumor removal may clarify the doubts regarding oncologic safety. Obviously, each patient should be approached individually with the appropriate surgical plan developed based on tumor type, location, and the clinical presentations. The goal of the surgery also will contribute to this decision making. Not all patients are suitable for endoscopic approaches. Small lesions and those that do not involve the carotid artery are more suitable for an exclusive endoscopic approach (108). Some patients may require combined approaches and in some, an open procedure may be beneficial. In one preliminary study, minimally invasive endoscopic approaches did not differ significantly from traditional craniofacial resection in operative time, blood loss, hospital stay or complication rate, survival, and recurrence rates (109). Larger prospective studies comparing both the techniques are required; the findings of such studies might help in proper selection of patients.
Tumor Resection
Figure 6 Basal view of the skull showing the extent of bone removal in fronto-orbital osteotomy. Source: From Ref. 89.
The idea behind the evolution of so many different cranial approaches has been to reduce brain retraction with improved visibility and accessibility of the tumor topography. Occasionally, however, it has been pointed out by us that extensive bony removal does not improve the visibility beyond a certain point and is thus unwarranted (110). Most extracranial tumors with intracranial extension would be exposed during the operative dissection itself, while they occupy the paranasal sinuses or the nasopharyngeal cavity. Tumors such
Chapter 19: Tumors of the Anterior Skull Base
as angiofibrolipoma, angiolipoma, and fibromatous tumors could be removed en bloc along with their attachment to the sinus mucosa. However, careful dissection is required in cases with dural attachment or infiltration. Discretion is required in radical resection of these tumors, keeping in mind that they could be invasive and malignant. In such instances, frozen section histopathology is helpful in decision making. Intracranial tumors with extracranial extension often require the detachment of the falx at the cribriform plate and division along the sagittal axis. Dural opening with adequate margin for closure allow for a good closure. The pterional approach with an orbital extension, would also afford an excellent view of planum sphenoidale and ACF when the Sylvian fissure is adequately split (100,111,112). The extent of craniotomy thus depends upon the size and location of the lesion as well as the clinical presentation of the patient (113). The principle behind safe removal of the intracranial tumor is maintaining the dissection within the arachnoid planes and avoiding pial breach. At places, this plane maybe eroded by the tumor, in which case, the pia has to be followed carefully to avoid injury to brain parenchyma and perforators from the anterior cerebral arteries (114). Cavitron ultrasonic aspirator has been a remarkable aid in achieving intratumoral resection and debulking. This also offers relaxation of the intracranial structures enabling clear visualization of neurovascular structures. Tumor debulking and periodical dissection around the tumor usually separates the surrounding structures away from the lesion. Attention is required in case of large tumors to save the perforators from the anterior communicating artery and the anterior cerebral arteries. In large tumors, these small vessels may be encased in the tumor or hidden on the other side of the tumor capsule. Endoscopy is an excellent accessory during surgery for large tumors that cross the midline. The bony part of the tumor or the hyperostotic bone in some meningiomas could pose difficulties. Most often, the exophytic part of the bone can be safely drilled away or completely removed if in close proximity to the craniotomy for the ease of subsequent reconstruction. Involved bone in the craniotomy flap should be excised and the bone flap reconstructed with methylmethacrylate or titanium mesh. Reconstruction En bloc removal of anterior skull base tumors may create defects in the anterior skull base and allow direct communication between the intracranial space and paranasal sinuses. Adequate reconstruction is essential to create a barrier between these two compartments, which will reduce the incidence of infection, CSF leak, and pneumocephalus. Small dural tears can be closed primarily or with fascia lata, temporalis fascia, or pericranium. Flaps for reconstruction of the anterior skull base can be classified as local flaps, pedicled myocutaneous flaps, and free tissue transfer. The most commonly used flap is the pericranial-galeal flap. The blood supply is derived from the supraorbital and supratrochlear arteries. These flaps are sufficient to repair small defects. If the orbital roofs are intact, the pericranial flap may be sufficient in achieving adequate support without formal bone reconstruction (115). Local flaps, like pericranial flaps, are difficult to harvest following previous surgery and prior radiotherapy. Galeal flaps may not be adequate to cover defects of bone like burr holes and craniotomy bony cuts. Loss of galea and frontalis muscle from the undersurface of the frontal scalp makes it thin and poorly vascularized and this might be problematic after radiotherapy (116,117). They can also cause functional and cosmetic com-
299
plications including sensory motor loss and regional alopecia (118). Medial transposition of the temporalis muscle is useful in reconstructing anterolateral skull base defects (115). The temporalis muscle flap derives its blood supply from the maxillary artery (119). In order to preserve the blood supply, the muscle should remain attached to the coronoid process (115). The temporalis muscle flap has a reliable blood supply while maintaining an effective arc of rotation of 120 to 130 degrees (120,121). Temporalis muscle can cover ipsilateral defects of the orbitofrontal region, maxilla, temporal bone, and infratemporal fossa. For reconstructing the orbit following exenteration, the temporalis muscle can be transposed into the orbit through a subcutaneous tunnel or through an opening in the lateral orbital wall (122). For maxillary reconstruction, the flap is transposed into the maxilla under the zygomatic arch. For more medially situated defects, temporalis muscle flaps are less reliable (115). Extensive skull base defects require either free tissue transfer from distant donor site or regionally transferred pedicled flaps. Regional myocutaneous flaps are an option in many skull base defects but free tissue transfer is more reliable (123). Regional pedicle flaps are often reserved for patients who are poor candidates for free flaps (124). The pectoralis major flap is widely used to repair extensive anterior and lateral skull base defects (124,125). Trapezius mycocutaneous flap is favored for reconstruction of lateral defects of the posterior skull base (115). Myocutaneous flaps may not always provide a watertight seal of the nasopharynx (126,127). Free tissue transfer is considered to be the best to repair extensive defects of the skull base by many authors (114,123,127–138). The success rates for free tissue transfer ranges from 86% to 100% (126,139–141). These free flap survival rates compare favorably to pedicled flaps used for similar conditions (142). The most common free flaps that are used in the reconstruction of anterior skull base defects are: rectus abdominis flap, the radial forearm flap, the latissimus dorsi flap, and the anterolateral thigh flap (123). Rectus abdominis is the most extensively used free flap for skull base reconstructions. It provides a bulky muscle that can obliterate dead space (115). Not only the defect size but individual characteristics of the defect and donor sites must also be taken into consideration when choosing the appropriate reconstruction source (123). For example, in obese patients, the anterolateral thigh flap is frequently used since it provides appropriate volume with a reliable skin paddle (143). Frequently used vessels for anastomoses include the facial, lingual arteries, and superior thyroid artery. The preferred recipient veins include the internal jugular, external jugular, and facial veins (123,137). End-to-end anastomoses are ordinarily performed with the exception of end-to-side anastomoses, which are performed when using the internal jugular vein (137). The incidence of complications following reconstruction of anterior skull base with free flaps is 10% to 42% (126,128,137,139,141,144,145). The risk of infection with flap reconstruction is approximately 1.4% to 7.4% (137). CSF leak following free flap transfer is about 0% to 8% (146). The advantage with free flap is the flexibility in flap content, and it is possible to introduce a large quantity of wellvascularized tissue in a single stage operation (117). High rate of complications are documented in elderly patients (147) in whom pedicled myocutaneous flaps are an alternative when large defects need reconstruction (124). The common complications are flap loss from ischemia and necrosis, flap failure resulting in CSF leak, meningitis, epidural abscess, empyema, and neurologic injury. Another concern is that the bulk of the
300
Javalkar et al.
free flap may mask local recurrence and radiologic follow-up is made more difficult (148). There are several options available for bone reconstruction. If the tumor removal requires craniotomy and the defect is small then there is no need for reconstruction. However, if the defect is large, this can be reconstructed with split calvarial bone graft, posterior wall of frontal sinus, or with materials like titanium mesh or methylmethacrylate (116). Some authors report that calvarial bone graft is unnecessary to repair large defects and combined galeal-pericranial flap is adequate for the reconstruction (149). In one study, authors have reported that large defects measuring 3.0 cm × 4.0 cm or greater must be reconstructed rigidly to prevent frontal lobe herniation into the paranasal sinuses (150). Some advocate a bone graft sandwiched between two pericranial flaps for reconstruction (151). Others advocate pericranial wrapping if postoperative radiation is planned (116), which might prevent osteoradionecrosis (152). If the medial–canthal ligaments are divided, then they need to be repositioned to the normal position. Nasolacrimal drainage needs to be re-established to avoid postoperative epiphora. Reconstruction of the medial wall of the orbit can be performed by using crossed sutures secured to the bone of the residual roof and floor of the orbit (88).
COMPLICATIONS AND AVOIDANCE The most common complications following craniofacial resections for anterior skull base tumors are infection, CSF leak, and pneumocephalus. In the International Study on Craniofacial Surgery, 33% of the patients experienced complications (77). The most common complications were related to the wound (18%) and central nervous system complications (14.8%). Mortality rate was 4.3%. In one study, it was found that 44% of patients may experience anosmia following anterior skull base surgery (117). Postoperative wound infection can lead to osteomyelitis of the bone flap. Intracranial extension can lead to meningitis and intracranial abscess formation. Care should be taken to meticulously repair the floor of the anterior cranial fossa to isolate the intracranial space from the paranasal sinuses. The frontal sinus should be cranialized and antibiotics used in the perioperative period. CSF leak can be avoided by watertight closure of the dural defect. The routine use of lumbar drain in all patients is controversial. Some advocate routine lumbar drain placement at the time of induction (75,87). Lumbar drain is not necessary in all cases, but is necessary if the dural closure is less than perfect. Pneumocephalus is one of the reasons for postoperative neurologic deterioration. This can be avoided by meticulous repair of the floor of the anterior cranial fossa to separate the cranial cavity from the nasal cavity and the paranasal sinuses. Patients should be instructed to avoid straining and blowing the nose in the postoperative period. The use of thrombin glue has been found to decrease pneumocephalus and air leakage (75).
POSTOPERATIVE CARE All patients with anterior skull base surgery require close observation in the intensive care setting. Any change in neurologic status warrants an urgent CT scan. Postoperative pneumocephalus is one of the causes for neurologic deterioration in the postoperative period following anterior skull base resection. In case of tension pneumocephalus, the trapped air
can be aspirated through an existing burr hole. If a spinal drain is in place, it needs to be closed. If this temporary measure fails, exploration of the craniotomy should be considered (112). Any new neurologic deficits are to be documented and evaluated with imaging. Postoperative MRI scans to document the extent of resection are routinely performed in the early postoperative period. Physical, occupational, and speech therapy are common postoperative treatments that these patients will require. In cases of skull base malignancies, postoperative adjuvant therapy like radiation or chemotherapy should be arranged. Postoperative radiation therapy is initiated within 4 to 6 weeks. Follow up appointments are scheduled on an individual basis. In case of anterior skull base malignancies, periodic examination by a team consisting of neurosurgeon, head and neck surgeon, radiation and medical oncologist are needed.
QUALITY OF LIFE FOLLOWING ANTERIOR SKULL BASE TUMORS There are not many studies in the literature that describe the quality of life following anterior skull base surgery (153,154). Gil et al. retrospectively evaluated quality of life following surgery for anterior skull base tumors (153). In their study, overall quality of life in most patients after anterior skull base tumor extirpation was good. Old age, malignancy, comorbidity, radiotherapy, and major surgery were found to be negative prognostic factors for quality of life measures. In a study by Fakuda et al., the authors evaluated the condition of 13 patients who underwent classic craniofacial resection of malignant tumors (154). Of the patients, 89% had some complaints and 63% of them reported that their quality of life worsened after surgery. Technical developments in anterior cranial fossa surgery has not only improved the long-term survival of patients but also improved the quality of life. REFERENCES 1. Dare AO, Balos LL, Grand W. Neural-dural transition at the medial anterior cranial base: An anatomical and histological study with clinical applications. J Neurosurg. 2003;99(2):362– 365. 2. Demonte F. Surgical treatment of anterior basal meningiomas. J Neurooncol. 1996;29:239–248. 3. Drummond KJ, Zhu JJ, Black PM. Meningiomas: Updating basic science, management, and outcome. Neurologist. 2004;10(3):113–130. 4. Berger L, Luc G, Richard D. The olfactory esthesioneuroepithelioma. Bull Assoc Franc Etude Cancer. 1924;13:410–421. 5. Schall LA, Linebeck M. Primary intranasal neuroblastoma: Report of 3 cases. Ann Otolaryngol. 1951;60:221–229. 6. McCormack LJ, Harris HE. Neurogenic tumors of the nasal fossa. JAMA. 1955;157:318–321. 7. Jethanamest D, Morris LG, Sikora AG, et al. Esthesioneuroblastoma: A population-based analysis of survival and prognostic factors. Arch Otolaryngol Head Neck Surg. 2007;133(3):276–280. 8. Resto VA, Eisele DW, Forastiere A, et al. Esthesioneuroblastoma: The Johns Hopkins experience. Head Neck. 2000;22:550–558. 9. Elkon D, Hightower SI, Linn ML, et al. Esthesioneuroblastoma. Cancer. 1979;44:1087–1094. 10. Spaulding CA, Kranyak MS, Constable WC, et al. Esthesioneuroblastoma: A comparison of two treatment eras. Int J Radiat Oncol Biol Phys. 1988;15:581–590. 11. Kadish S, Goodman M, Wang CC. Olfactory neuroblastoma: A clinical analysis on 17 cases. Cancer. 1976;37:1571–1576. 12. McLean JN, Nunley SR, Klass C, et al. Combined modality therapy of esthesioneuroblastoma. Otolaryngol Head Neck Surg. 2007;136(6):998–1002.
Chapter 19: Tumors of the Anterior Skull Base 13. Eich HT, Muller RP, Micke O, et al. Esthesioneuroblastoma in childhood and adolescence: Better prognosis with multimodal treatment? Strahlenther Onkol. 2005;181(6):378–384. 14. Cianfriglia F, Pompili A, Occhipinti E. Intracranial malignant cartilaginous tumours. Report of two cases and review of literature. Acta Neurochir (Wien). 1978;45:163–175. 15. Brackmann DE, Teufert KB. Chondrosarcomas of the skull base: Long-term follow-up. Otol Neurotol. 2006;27:981–991. 16. Barnes L, Kapadia SB. The biology and pathology of selected skull base tumors. J Neurooncol. 1994;20:213–240. 17. Brooks JJ, LiVolsi VA, Trojanowski JQ. Does chondroid chordoma exist? Acta Neuropathol. 1987;72:229–235. 18. Richardson MS. Pathology of skull base tumors. Otolaryngol Clin North Am. 2001;34:1025–1042. 19. Rosenberg AE, Nielsen GP, Keel SB, et al. Chondrosarcoma of the base of the skull: A clinicopathologic study of 200 cases with emphasis on its distinction from chordoma. Am J Surg Pathol. 1999;23:1370–1378. 20. La Rocca RV, Morgan KW, Paris K, et al. Recurrent chondrosarcoma of the cranial base: A durable response to ifosfamidedoxorubicin chemotherapy. J Neurooncol. 1999;41:281–283. 21. Marshman LA, Gunasekera L, Rose PE, et al. Primary intracerebral mesenchymal chondrosarcoma with rhabdomyosarcomatous differentiation: Case report and literature review. Br J Neurosurg. 2001;15:419–424. 22. Tzortzidis F, Elahi F, Wright D, et al. Patient outcome at long-term follow-up after aggressive microsurgical resection of cranial base chondrosarcomas. Neurosurgery. 2006;58(6):1090– 1098. 23. Heffelfinger MJ, Dahlin DC, MacCarty CS, et al. Chordomas and cartilaginous tumors at the skull base. Cancer. 1973;32:410–420. 24. Hoch BL, Nielsen GP, Liebsch NJ, et al. Base of skull chordomas in children and adolescents: A clinicopathologic study of 73 cases. Am J Surg Pathol. 2006;30(7):811–818. 25. Gay E, Sekhar LN, Rubinstein E, et al. Chordomas and chondrosarcomas of the cranial base: Results and follow-up of 60 patients. Neurosurgery. 1995;36:887–897. 26. Raffel C, Wright DC, Gutin PH, et al. Cranial chordomas: Clinical presentation and results of operative and radiation therapy in 26 patients. Neurosurgery. 1985;17:703–710. 27. Meis JM, Raymond AK, Evans HL, et al. Dedifferentiated chordoma. A clinicopathologic and immunohistochemical study of three cases. Am J Surg Pathol. 1987;11(7):516–525. 28. Hruban RH, Traganos F, Reuter VE, et al. Chordomas with malignant spindle cell components. A DNA flow cytometric and immunohistochemical study with histogenetic implications. Am J Pathol. 1990;137(2):435–447. 29. Casali PG, Stacchiotti S, Sangalli C, et al. Chordoma. Curr Opin Oncol. 2007;19(4):367–370. 30. Casali PG, Messina A, Stacchiotti S, et al. Imatinib mesylate in chordoma. Cancer. 2004;101:2086–2097. 31. Wei WI, Sham JS. Nasopharyngeal carcinoma. Lancet. 2005;365(9476):2041–2054. 32. Gulley ML. Molecular diagnosis of Epstein-Barr virus-related diseases. J Mol Diagn. 2001;3(1):1–10. 33. Shanmugaratnam K, Sobin LH. Histological typing of tumors of the upper respiratory tract and ear. In: Shanmugaratnam K, Sobin LH, eds. International Histological Classification of Tumors: No 19. Geneva: WHO, 1991:32–33. 34. ESMO Guidelines working group, Chan AT, Hui EP, Leung SF. Nasopharyngeal cancer. ESMO clinical recommendations for diagnosis, treatment and follow-up. Ann Oncol. 2007;18(Suppl 2):ii67–ii68. 35. Li JH, Chia M, Shi W, et al. Tumor-targeted gene therapy for nasopharyngeal carcinoma. Cancer Res. 2002;62(1):171–178. 36. Chua D, Huang J, Zheng B, et al. Adoptive transfer of autologous Epstein-Barr virus-specific cytotoxic T cells for nasopharyngeal carcinoma. Int J Cancer. 2001;94(1):73–80. 37. Chang AE, Karnell LH, Menck HR. The national database report on cutaneous and noncutaneous melanomas: a summary of 84,836 cases from the past decade. The American College of Surgeons Commission of Cancer and the American Society. Cancer. 1998;83:1664–1678.
301
38. Ganly I, Patel SG, Singh B, et al. Craniofacial resection for malignant melanoma of the skull base: Report of an international collaborative study. Arch Otolaryngol Head Neck Surg. 2006;132(1):73–78. 39. Vrionis FD, Kienstra MA, Rivera M, et al. Malignant tumors of the anterior skull base. Cancer Control. 2004;11(3):144–151. 40. Abbondanzo SL, Wenig BM. Non-Hodgkin’s lymphoma of the sinonasal tract. A clinicopathologic and immunophenotypic study of 120 cases. Cancer. 1995;75:1281–1291. 41. Shohat I, Berkowicz M, Dori S, et al. Primary non-Hodgkin’s lymphoma of the sinonasal tract. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2004;97(3):328–331. 42. Kaplan MJ, Fischbein NJ, Harsh GR. Anterior skull base surgery. Otolaryngol Clin North Am. 2005;38(1):107–131. 43. Dare AO, Datta RV, Loree TR, et al. Sinonasal nonHodgkin’s lymphoma with skull base involvement. Skull Base. 2001;11(2):129–135. 44. Danks RA, Kaye AH. Carcinoma of the paranasal sinuses. In: Kaye A, Laws ER Jr, eds. Brain Tumours: An Encyclopedic Approach. New York, NY: Churchill Livingstone, 1995:809–824. 45. Shima N, Kobashi Y, Tsutsia K, et al. Extranodal non-Hodgkins lymphoma of the head and neck. A clinico-pathologic study in the Kyoto-Nara area in Japan. Cancer. 1990;66:1190–1197. 46. Quraishi MS, Bessell EM, Clark D, et al. Non-Hodgkin’s lymphoma of the sinonasal tract. Laryngoscope. 2000;110:1489– 1492. 47. Vidal RW, Devaney K, Ferlito A, et al. Sinonasal malignant lymphomas: A distinct clinicopathological category. Ann Otol Rhinol Laryngol. 1999;108:411–419. 48. Logsdon M, Ha CS, Kavadi VS, et al. Lymphoma of the nasal cavity and paranasal sinuses: Improved outcome and altered prognostic factors with combined modality therapy. Cancer. 1997;80:477–488. 49. Bales C, Kotapka M, Loevner LA, et al. Craniofacial resection of advanced juvenile nasopharyngeal angiofibroma. Arch Otolaryngol Head Neck Surg. 2002;128(9):1071–1078. 50. Cummings BJ, Blend R, Keane T, et al. Primary radiation therapy for juvenile nasopharyngeal angiofibroma. Laryngoscope. 1984;94:1599–1605. ¨ 51. Issing PR, Hemmanouil I, Stover T, et al. Adenoid cystic carcinoma of the skull base. Skull Base Surg. 1999;9(4):271–275. 52. Nascimento AG, Amaral AL, Prado LA, et al. Adenoid cystic carcinoma of salivary glands. A study of 61 cases with clinicopathologic correlation. Cancer. 1986;57(2):312–319. 53. Pitman KT, Prokopakis EP, Aydogan B, et al. The role of skull base surgery for the treatment of adenoid cystic carcinoma of the sinonasal tract. Head Neck. 1999;21(5):402–407. 54. Matsuba HM, Spector GJ, Thawley SE, et al. Adenoid cystic salivary gland carcinoma: A histologic review of treatment failure patterns. Cancer. 1986;57:519–524. 55. Pommier P, Liebsch NJ, Deschler DG, et al. Proton beam radiation therapy for skull base adenoid cystic carcinoma. Arch Otolaryngol Head Neck Surg. 2006;132(11):1242–1249. 56. Bradley PJ. Adenoid cystic carcinoma of the head and neck: A review. Curr Opin Otolaryngol Head Neck Surg. 2004;12(2):127– 132. 57. Mori Y, Kobayashi T, Kida Y, et al. Stereotactic radiosurgery as a salvage treatment for recurrent skull base adenoid cystic carcinoma. Stereotact Funct Neurosurg. 2005;83(5–6):202–207. 58. Lee N, Millender LE, Larson DA, et al. Gamma knife radiosurgery for recurrent salivary gland malignancies involving the base of skull. Head Neck. 2003;25(3):210–216. 59. Pieper DR, LaRouere M, Jackson IT. Operative management of skull base malignancies: Choosing the appropriate approach. Neurosurg Focus. 2002;12(5):e6. 60. Blandino A, Gaeta M, Minutoli F, et al. CT and MR findings in neoplastic perineurla spread along the vidian nerve. Eur Radiol. 2000;10:521–526. 61. Laine FJ, Braun IF, Jensen ME, et al. Perineural tumor extension through the foramen ovale: Evaluation with MR imaging. Radiology. 1990;174(1):65–71. 62. Hanna E, Vural E, Prokopakis E, et al. The sensitivity and specificity of high-resolution imaging in evaluating perineural
302
63.
64.
65.
66.
67. 68. 69. 70.
71.
72.
73.
74.
75. 76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
Javalkar et al. spread of adenoid cystic carcinoma to the skull base. Arch Otolaryngol Head Neck Surg. 2007;133(6):541–545. Connor SE, Umaria N, Chavda SV. Imaging of giant tumours involving the anterior skull base. Br J Radiol. 2001;74(883):662– 667. Yousem DM, Gad K, Tufano RP. Resectability issues with head and neck cancer. AJNR Am J Neuroradiol. 2006;27(10):2024– 2036. Yousem DM, Hatabu H, Hurst RW, et al. Carotid artery invasion by head and neck masses: Prediction with MR imaging. Radiology. 1995;195:715–720. Yu ZH, Xu GZ, Huang YR, et al. Value of computed tomography in staging the primary lesion (T-staging) of nasopharyngeal carcinoma (NPC): An analysis of 54 patients with special reference to the parapharyngeal space. Int J Radiat Oncol Biol Phys. 1985;11:2143–2147. Ginsberg LE. MR imaging of perineural tumor spread. Neuroimaging Clin N Am. 2004;14:663–677. Ginsberg LE. MR imaging of perineural tumor spread. Magn Reson Imaging Clin N Am. 2002;10:511–525. Ginsberg LE. Imaging of perineural tumor spread in head and neck cancer. Semin Ultrasound CT MR. 1999;20:175–186. Babin E, Hamon M, Benateau H, et al. Interest of PET/CT scan fusion to assess mandible involvement in oral cavity and oropharyngeal carcinomas [in French]. Ann Otolaryngol Chir Cervicofac. 2004;121:235–240. Gil Z, Even-Sapir E, Margalit N, et al. Integrated PET/CT system for staging and surveillance of skull base tumors. Head Neck. 2007;29(6):537–545. Schwartz DL, Ford E, Rajendran J, et al. FDG-PET/CT imaging for preradiotherapy staging of head-and-neck squamous cell carcinoma. Int J Radiat Oncol Biol Phys. 2005;61:129–136. Yeung HW, Schoder H, Smith A, et al. Clinical value of combined positron emission tomography/computed tomography imaging in the interpretation of 2-deoxy-2-[F-18]fluoro-d-glucosepositron emission tomography studies in cancer patients. Mol Imaging Biol. 2005;7:229–235. Fukui MB, Blodgett TM, Meltzer CC. PET/CT imaging in recurrent head and neck cancer. Semin Ultrasound CT MR. 2003;24:157–163. Shah JP, Bilsky MH, Patel SG. Malignant tumors of the skull base. Neurosurg Focus. 2002;13(4):e6. Cantu` G, Riccio S, Bimbi G, et al. Craniofacial resection for malignant tumours involving the anterior skull base. Eur Arch Otorhinolaryngol. 2006;263(7):647–652. Patel SG, Singh B, Polluri A, et al. Craniofacial surgery for malignant skull base tumors: Report of an international collaborative study. Cancer. 2003;98(6):1179–1187. Smith RR, Klopp CT, Williams JM. Surgical treatment of cancer of the frontal sinus and adjacent areas. Cancer. 1954;7(5):991– 994. Ketcham AS, Hoye RC, Van Buren JM, et al. Complications of intracranial facial resection for tumors of the paranasal sinuses. Am J Surg. 1966;112(4):591–596. Shah JP, Krauss DH, Bilsky MH, et al. Craniofacial resection for malignant tumors involving the anterior skull base. Arch Otolaryngol Head Neck Surg. 1997;123:1312–1317. Suarez C, Llorente JL, Fernandez De Leon R, et al. Prognostic factors in sinonasal tumors involving the anterior skull base. Head Neck. 2004;26(2):136–144. Kraus DH, Gonen M, Mener D, et al. A standardized regimen of antibiotics prevents infectious complications in skull base surgery. Laryngoscope. 2005;115(8):1347–1357. Robbins KT, Byers RM, Cole R, et al. Wound prophylaxis with metronidazole in head and neck surgical oncology. Laryngoscope. 1988;98:803–806. Carrau RL, Snyderman C, Janecka IP, et al. Antibiotic prophylaxis in cranial base surgery. Head Neck. 1991;13(4):311– 317. ´ Rodrigo JP, Alvarez JC, Gomez JR, et al. Comparison of three prophylactic antibiotic regimens in clean-contaminated head and neck surgery. Head Neck. 1997;19(3):188–193.
86. Reinhart E, Reuther J, Michel C, et al. Perioperative antibiotic prophylaxis in orthodontic bone operations of the facial skull. Mund Kiefer Gesichtschir [in German]. 1998;2(4):194– 201. 87. Wong LY, Lam LK, Fan YW, et al. Outcome analysis of patients with craniofacial resection: Hong Kong experience. ANZ J Surg. 2006;76(5):313–317. 88. Solero CL, DiMeco F, Sampath P, et al. Combined anterior craniofacial resection for tumors involving the cribriform plate: Early post-operative complications and technical considerations. Neurosurgery. 2000;47(6):1296–1305. 89. Acharya R, Shaya M, Kumar R, et al. Quantification of the advantages of the extended frontal approach to skull base. Skull Base. 2004;14(3):133–142. 90. Shah JP, Bilsky MH, Patel SG. Malignant tumors of the skull base. Neurosurg Focus. 2002;13(4):e6. 91. Raveh J, Vuillemin T, Sutter F. Subcranial management of 395 combined frontobasal-midface fractures. Arch Otolaryngol Head Neck Surg. 1988;114(10):1114–1122. 92. Raveh J, Turk JB, L¨adrach K, et al. Extended anterior subcranial approach for skull base tumors: Long-term results. J Neurosurg. 1995;82(6):1002–1010. 93. Kinnunen I, Aitasalo K. A review of 59 consecutive patients with lesions of the anterior cranial base operated on using the subcranial approach. J Craniomaxillofac Surg. 2006;34(7):405– 411. 94. Har-El G, Lucente FE. Midfacial degloving approach to the nose, sinuses and skull base. Am J Rhinol. 1996;10:17–22. 95. Cocke EW, Robertson JH. Extended unilateral maxillotomy approach. In: Donald PJ, ed. Surgery of the Skull Base. Philadelphia, PA: Lippincott-Raven, 1998:207–237. 96. Pellerin P, Lesoin F, Dhellemmes P, et al. Usefulness of the orbitofrontomalar approach associated with bone reconstruction for frontotemporosphenoid meningiomas. Neurosurgery. 1984;15(5):715–718. 97. Hakuba A, Liu S, Nishimura S. The orbitozygomatic infratemporal approach: A new surgical technique. Surg Neurol. 1986;26(3):271–276. 98. Hendryk S, Czecior E, Misiołek M, et al. Surgical strategies in the removal of malignant tumors and benign lesions of the anterior skull base. Neurosurg Rev. 2004;27(3):205–213. 99. Hassler W, Zentner J. Pterional approach for surgical treatment of olfactory groove meningiomas. Neurosurgery. 1989;25:942– 947. 100. Sekhar LN, Nanda A, Sen CN, et al. The extended frontal approach to tumors of the anterior, middle and posterior skull base. J Neurosurg. 1992;76:198–206. 101. Otani N, Muroi C, Yano H, et al. Surgical management of tuberculum sellae meningioma; role of selective extradural anterior clinoidectomy. Br J Neurosurg. 2006;20(3):129–138. 102. Arai H, Sato K, Okuda, et al. Transcranial transsphenoidal approach for tuberculum sellae meningiomas. Acta Neurochir (Wien). 2001;142(7):751–757. 103. Hallacq P, Moreau JJ, Fischer G, et al. Frontal sinus approach to olfactory groove meningiomas [in French]. Neurochirurgie. 1999;45(4):329–337. 104. Jho HD. Orbital roof craniotomy via an eyebrow incision: A simplified anterior skull base approach. Minim Invasive Neurosurg. 1997;40(3):91–97. 105. Jho HD, Ha HG. Endoscopic endonasal skull base surgery: Part 1 - The midline anterior fossa skull base. Minim Invasive Neurosurg. 2004;47:1–8. 106. Jho HD, Alfieri A. Endoscopic glabellar approach to the anterior skull base: A technical note. Minim Invasive Neurosurg. 2002;45(3):185–188. 107. Casiano RR, Numa WA, Falquez AM. Endoscopic resection of esthesioneuroblastoma. Am J Rhinol. 2001;15:271–279. 108. Casler JD, Doolittle AM, Mair EA. Endoscopic surgery of the anterior skull base. Laryngoscope. 2005;115(1):16–24. 109. Batra PS, Citardi MJ, Worley S, et al. Resection of anterior skull base tumors: Comparison of combined traditional and endoscopic techniques. Am J Rhinol. 2005;19(5):521–528.
Chapter 19: Tumors of the Anterior Skull Base 110. Nanda A, Vincent DA, Vannemreddy P, et al. Far-lateral approach to intradural lesions of the foramen magnum without resection of the occipital condyle. J Neurosurg. 2002;96(2):302– 309. 111. Delashaw JB Jr, Tedeschi H, Rhoton AL. Modified supraorbital craniotomy: Technical note. Neurosurgery. 1992;30(6):954–956. 112. Kiyokawa K, Tai Y, Yanaga H, et al. Evaluation with threedimensional computed tomography after anterior skull base reconstruction using two musculopericranial flaps and a grafted bone. Skull Base Surg. 1999;9(3):221–226. 113. Spektor S, Valarezo J, Fliss DM, et al. Olfactory groove meningiomas from neurosurgical and ear, nose, and throat perspectives: Approaches, techniques, and outcomes. Neurosurgery. 2005;57(4 Suppl):268–280; discussion 268–280. 114. Thomson JG, Restifo RJ. Microsurgery for cranial base tumors. Clin Plast Surg. 1995;22:563–572. 115. Liu JK, Niazi Z, Couldwell WT. Reconstruction of the skull base after tumor resection: An overview of methods. Neurosurg Focus. 2002;12(5):e9. 116. Gil Z, Abergel A, Leider-Trejo L, et al. A comprehensive algorithm for anterior skull base reconstruction after oncological resections. Skull Base. 2007;17(1):25–37. 117. Fliss DM, Gil Z, Spektor S, et al. Skull base reconstruction after anterior subcranial tumor resection. Neurosurg Focus. 2002;12(5):e10. 118. Snyderman CH, Janecka IP, Sekhar LN, et al. Anterior cranial base reconstruction: Role of galeal and pericranial flaps. Laryngoscope. 1990;100(6):607–614. 119. Elazab EE, Abdel-Hameed FA. The arterial supply of the temporalis muscle. Surg Radiol Anat. 2006;28(3):241–247. 120. Bradley P, Brockbank J. The temporalis muscle flap in oral reconstruction. A Cadaveric, animal and clinical study. J Maxillofac Surg. 1981;9:139–145. 121. Birt BD, Antonyshyn O, Gruss JS. The temporalis muscle flap for head and neck reconstruction. J Otolaryngol. 1987;16:179–118. 122. Clauser L, Curioni C, Spanio S. The use of the temporalis muscle flap in facial and craniofacial reconstructive surgery. A review of 182 cases. J Craniomaxillofac Surg. 1995;23:203–214. 123. Pusic AL, Chen CM, Patel S, et al. Microvascular reconstruction of the skull base: A clinical approach to surgical defect classification and flap selection. Skull Base. 2007;17(1):5–15. 124. Resto VA, McKenna MJ, Deschler DG. Pectoralis major flap in composite lateral skull base defect reconstruction. Arch Otolaryngol Head Neck Surg. 2007;133(5):490–494. 125. Sasaki CT, Ariyan S, Spencer D, et al. Pectoralis major myocutaneous reconstruction of the anterior skull base. Laryngoscope. 1985;95(2):162–166. 126. Neligan PC, Mulholland S, Irish J, et al. Flap selection in cranial base reconstruction. Plast Reconstr Surg. 1996;98(7):1159–1166. 127. Yamada A, Harii K, Ueda K, et al. Free rectus abdominis muscle reconstruction of the anterior skull base. Br J Plast Surg. 1992;45(4):302–306. 128. Demonte F, Moore BA, Chang DW. Skull base reconstruction in the pediatric patient. Skull Base. 2007;17(1):39–51. 129. Weber SM, Kim JH, Wax MK. Role of free tissue transfer in skull base reconstruction. Otolaryngol Head Neck Surg. 2007;136(6):914–919. 130. Djalilian HR, Gapany M, Levine SC. Reconstruction of complicated skull base defects utilizing free tissue transfer. Skull Base. 2002;12(4):209–213. 131. Hurvitz KA, Kobayashi M, Evans GR. Current options in head and neck reconstruction. Plast Reconstr Surg. 2006;118(5):122e– 133e. 132. Disa JJ, Rodriguez VM, Cordeiro PG. Reconstruction of lateral skull base oncological defects: The role of free tissue transfer. Ann Plast Surg. 1998;41:633–639. 133. Neligan PC, Boyd JB. Reconstruction of the cranial base defect. Clin Plast Surg. 1995;22:71–77.
303
134. Jones TR, Jones NF. Advances in reconstruction of the upper aerodigestive tract and cranial base with free tissue transfer. Clin Plast Surg. 1992;19:819–831. 135. Chang DW, Robb GL. Microvascular reconstruction of the skull base. Semin Surg Oncol. 2000;19:211–217. 136. Valentini V, Fabiani F, Nicolai G, et al. Use of microvascular free flaps in the reconstruction of the anterior and middle skull base. J Craniofac Surg. 2006;17(4):790–796. 137. Califano J, Cordeiro PG, Disa JJ, et al. Anterior cranial base reconstruction using free tissue transfer: Changing trends. Head Neck. 2003;25(2):89–96. 138. Besteiro JM, Aki FE, Ferreira MC, et al. Free flap reconstruction of tumors involving the cranial base. Microsurgery. 1994;15:9– 13. 139. Urken ML, Catalano PJ, Sen C, et al. Free tissue transfer for skull base reconstruction analysis of complications and a classification scheme for defining skull base defects. Arch Otolaryngol Head Neck Surg. 1993;119:1318–1325. 140. Jones NF, Sekhar LN, Schramm VL. Free rectus abdominis muscle flap reconstruction of the middle and posterior cranial base. Plast Reconstr Surg. 1986;78:471–479. 141. Clayman GL, DeMonte F, Jaffe DM, et al. Outcome and complications of extended cranial-base resection requiring microvascular free-tissue transfer. Arch Otolaryngol Head Neck Surg. 1995;121:1253–1257. 142. Newman J, O’Malley BW, Chalian A, et al. Microvascular reconstruction of cranial base defects: An evaluation of complication and survival rates to justify the use of this repair. Arch Otolaryngol Head Neck Surg. 2006;132(4):381–384. 143. Davidge K, Pusic A, Disa JJ, et al. Use of the anterolateral thigh flap as an alternative to the rectus flap in obese and over-weight patients. Ann Plast Surg. 2006;56:536–539. 144. Izquierdo R, Leonetti JP, Origitano TC, et al. Refinements using free-tissue transfer for complex cranial base reconstruction. Plast Reconstr Surg. 1993;92:567–574. 145. Chang DW, Langstein HN, Gupta A, et al. Reconstructive management of cranial base defects after tumor ablation. Plast Reconstr Surg. 2001;107(6):1346–1357. 146. Weber SM, Kim JH, Wax MK. Role of free tissue transfer in skull base reconstruction. Otolaryngol Head Neck Surg. 2007;136(6):914–919. 147. Coleman JJ. Microvascular approach to function and appearance of large orbital maxillary defects. Am J Surg. 1989;158(4):337–341. 148. Kiyokawa K, Tai Y, Inoue Y, et al. Efficacy of temporal musculopericranial flap for reconstruction of the anterior base of the skull. Scand J Plast Reconstr Surg Hand Surg. 2000;34(1):43–53. 149. Abe T, Goda M, Kamida T, et al. Overlapping free bone graft with galeal-pericranium in reconstruction of the anterior skull base prevented CSF leak and sequestrum formation. Acta Neurochir (Wien). 2007;149:771–775. 150. Rodrigues M, O’malley BW, Staecker H, et al. Extended pericranial flap and bone graft reconstruction in anterior skull base surgery. Otolaryngol Head Neck Surg. 2004;131:69–76. 151. Gil Z, Fliss DM. Pericranial wrapping of the frontal bone after anterior skull base tumor resection. Plast Reconstr Surg. 2005;116:395–398. 152. Origitano TC, Petruzzelli GJ, Leonetti JP, et al. Combined anterior and anterolateral approaches to the cranial base: Complication analysis, avoidance, and management. Neurosurgery. 2006;58(4 Suppl 2):ONS-327–336. 153. Gil Z, Abergel A, Spektor S, et al. Quality of life following surgery for anterior skull base tumors. Arch Otolaryngol Head Neck Surg. 2003;129(12):1303–1309. 154. Fukuda K, Saeki N, Mine S, et al. Evaluation of outcome and QOL in patients with craniofacial resection for malignant tumors involving the anterior skull base. Neurol Res. 2000;22:545– 550.
20 Infratemporal/Middle Fossa Tumors Paul J. Donald
cavity and the upper aerodigestive system without undue hazard to the patient. The outcomes of the studies by Ketcham et al. (3, 4) and Sisson et al. (5) in the resection of extensive paranasal sinus tumors proved that combined craniofacial surgery can be done safely and with a low infection rate. The cooperation between the neurologic surgeon and the head and neck surgeon should be “natural,” since their respective skills are complementary with regard to the skull base region. The areas that hold a morbid fear for the head and neck surgeon—the contents of the cranial cavity—are regions of easy familiarity for the neurosurgeon. Those areas beneath the skull base and deep in the face that for the neurosurgeon are an anatomic and physiologic mystery are common ground for his cohort in the head and neck. The advances in modern anesthesia are largely responsible for the safety of the technical manipulations of the surgical team. Air embolism is a constant threat, especially if the patient is operated upon in the sitting position. In general, most of the procedures are performed in the supine position. However, despite this, air embolism via the dural venous sinuses or the large veins in the neck, although less frequently a problem, is still a constant threat. Considerable evidence has been accumulated to show that the Doppler ultrasound can monitor the emergence of an air embolus much more precisely than any other monitoring system (6). Its sensitivity compared to other methods is illustrated in Table 1. In a dog, a volume of air as scant as 0.12 to 0.25 mL can be detected. The characteristic murmur of an air embolus heard over the precordium is not heard until at least 15 to 30 mL of air is present (7). Monitoring of intracranial pressure with a catheter in the third ventricle permits an indirect assessment of intracranial volume. A clear understanding of the relationship between intracranial volume and intracranial pressure is vital. Diminution of intracranial volume is an essential step in promoting optimal conditions for the neurosurgical part of the operation. A number of measurements are involved in determining this critical factor. As cerebral blood flow increases, so does the cerebral volume (Fig. 1). A mean circulating blood pressure between 60 and 100 mm Hg has little effect on the blood flow to the brain. However, once 100 mm Hg is exceeded, cerebral blood flow and concomitantly, intracranial volume increase. Cerebral vascular tone is exquisitely sensitive to concentrations of PCO2 . An elevation of PCO2 of 1 mm Hg will increase the cerebral blood flow by 2%. Although intracranial blood flow, and hence intracranial volume, is highly sensitive to PCO2 , there is no change in cerebral blood flow when the PaO2 is between 40 and 300 mm Hg. Therefore it is essential to maintain a low PCO2 during the intracranial part of the procedure. The insertion of a Dean tube, a catheter inserted into the endotracheal tube and connected to a CO2 analyzer, records end tidal CO2 , providing a continuous precise monitoring of this vital parameter. Care
Possibly the most challenging area in head and neck surgery is the region of the skull base. It was only a short time ago that neoplasms, especially malignancies close or adherent to the undersurface of the cranium were considered inoperable. Once the barrier had been breached and tumor had spread intracranially, the patient’s prognosis became entirely hopeless. This pessimistic situation has radically changed in recent years, however, thanks to the concurrence of three major developments: the advent of microvascular surgery, the emergence of the era of combined craniofacial surgery, and the advances in modern anesthesia in both the areas of pharmacologic manipulation and monitoring. In 1972, McLean and Buncke (1) first transplanted omentum to the cranium using a microvascular anastomosis. During the past decade, the work of such pioneers as O’Brien, Acland, Taylor, and Harii established the techniques of free flap transplantation as a viable clinical tool. The application of these principles to the revascularization of compromised cerebral circulation was one of the major breakthroughs in expanding the scope and safety of skull base surgery (2). One of the principal stumbling blocks in this area has always been the inviolability of the internal carotid artery. If this vessel was compromised, the viability of the ipsilateral cerebral hemisphere became severely jeopardized. Despite the assurance of numerous investigative procedures indicating collateral supply, this viability could not be guaranteed. The development of the superficial, temporal-to-middle cerebral artery vascular anastomosis had hopes of making internal carotid artery sacrifice possible (3). However, the low flow often seen through this anastomosis may be insufficient to provide adequate supply to the hemisphere deprived of its internal carotid supply. Currently, the internal carotid artery is always replaced when feasible even when patients pass their balloon test occlusion study and SPECT scan (see later). In recent years, much has been made of the so-called “team approach,” in which paramedical and medical personnel coordinate their efforts in the management of patients. The impetus for this has largely come from the paramedical disciplines. It is curious that this most laudable concept had somewhat slow acceptance among some of the surgical disciplines when it concerned cooperation between certain surgical specialties. Cooperation between head and neck surgeons, especially otorhinolaryngologists and neurologic surgeons was initially halting. The neurosurgeons’ reluctance to cooperate was based not so much on a lack of trust in their counterparts’ ability to handle the cervicofacial end of the surgery, but on a deep concern that the carefully guarded, sterile cranial cavity would be exposed to the profusion of microorganisms inhabiting the environs of the upper aerodigestive tract. The work of Paul Tessier, probably more than anyone, clearly demonstrated that a substantial surgical communication could be temporarily created between the cranial 305
306
Donald
Table 1 Detectable Levels of Emboli in Dogs Method Doppler Decreased end-tidal CO2 Elevated pulmonary artery pressure Decreased blood pressure Electrocardiography
Embolus volume per body weight (mL/kg) 0.01–0.02 0.10 0.25–0.50 0.50–1.00 0.25–0.50 4.00–8.00 4.00–8.00
Source: From Ref. 67.
must be taken to maintain the airway pressure on the respirator below 25 cm H2 O. Pressures in excess of this will raise intracranial pressure. Anesthetic agents may have an effect on cerebral blood flow as well. Inducting agents such as Althesin and thiopental (Pentothal) bring about a decrease in cerebral flow, and hence a fall in intracranial pressure. On the other hand, halothane causes an increase in pressure. Although methoxyflurane (Penthrane) appears to be the agent of choice at some institutions, (8) the best anesthetic for minimizing cerebral flow is a combination of nitrous oxide and narcotics (Table 2). The profusion of blood vessels in the region of the skull base, as well as in the equivalent area intracranially, creates a problem of significant blood loss. Controlled hypotension using nitroprusside or similar agents reduces hemorrhage markedly. A mean circulating blood pressure of 60 to 70 mm Hg adequately perfuses vital organs while markedly facilitating dissection, thus reducing both blood loss and operating time. Central venous and arterial pressure monitoring devices are essential in the safe application of this technique. A continuous digital readout displaying these parameters, as well as pulse and respiratory rates, greatly facilitates observation of the patient’s intraoperative course. Success of a combined intracranial/extracranial surgery depends not only on the smooth coordination of the surgical team but on expert nursing care as well. Diligent postoperative observation and rapid intervention in the event of an alteration in the patient’s status is vital in the care of these patients. Following a technically successful procedure, the loss of life or the development of permanent morbid sequelae as a result of inadequate, inattentive postoperative care is an execrable tragedy. A therapeutic dilemma that occasionally follows extensive craniofacial procedures is the problem of poor systemic perfusion in a hypotensive patient in whom restoration of an adequate blood pressure with intravascular volume augmentation will result in an intolerable increase in cerebral volume. Early intervention by an experienced, skilled medical team for the delicate management of volume replacement and pharmacologic regulation of systemic pressure is the only hope of saving the patient from the ravages of profound cerebral edema (Fig. 1). Monitoring of the patient’s hemodynamics, blood gases, and cerebrospinal fluid pressure must be continuous and compulsively exercised pursuits.
ANATOMY Skull base anatomy is complex, requiring the head and neck surgeon who endeavors to perform surgical procedures here to have a clear and precise three-dimensional conception of the topography and contents of the region. Moreover, it is mandatory to comprehend the structures that lie on both sides of the cranial surfaces of the plates and buttresses on
Figure 1 A pressure–volume curve relating rises of intracranial pressure to increases in intracerebral volume. Pressure is in millimeters of Hg and volume is in cubic centimeters. This curve was extrapolated from an experiment relating intracranial volume and pressure using an inflatable balloon in the cranium of a Rhesus monkey. Source: From Ref. 53.
which one is working. Much like acquiring an appreciation of the anatomy of the temporal bone during dissections prior to performing ear surgery, the central anatomy must be studied “from the outside in”. Preparation for surgery is done in the dissecting laboratory or the autopsy room where, with the aid of anatomic texts, a spatial conception of this region can be formulated. Surgery in the skull base area is very demanding but is not outside the scope of any well-trained head and neck surgeon who is willing to do a little homework and acquire the necessary team members. Nasopharyngeal carcinoma has been generally considered and still remains essentially primarily a nonsurgical disease. The mucosa of this region suffers the disadvantage of being adherent and adjacent to a number of vital structures. It is tightly affixed to the bone of the clivus, precluding a clean plane of dissection, as is usually found throughout the rest of the pharynx. In addition, the internal carotid artery in the lateral recesses is separated from the nasopharynx by a mere plug of cartilage in the inferior aspect of the foramen lacerum, and the brain stem is located on the deep surface of the clivus (Fig. 2). However, despite these complicating factors, recent surgical advances have permitted nasopharyngeal resections in patients who have failed combined radiation and chemotherapy. The Eustachian tube enters the region of the nasopharynx through the foramen of Morgagni formed by the upper portion of the superior constrictor muscle that forms a sort of sling under the tube. The superior constrictor muscle of the nasopharynx has its origin on the pharyngeal tubercles of the basiocciput and inserts into the inferior aspect of the medial pterygoid plate and hamulus as well as the pterygomandibular raphe. The tubercles are the origin of both the pharyngobasilar and the buccopharyngeal fascial sheaths that encase the superior constrictor muscle in its anterior and posterior aspects respectively (Fig. 3). The remainder of the clivus is covered by a thick layer of mucosa that overlies a very heavy layer
Chapter 20: Infratemporal/Middle Fossa Tumors
307
Figure 2 Coronal section through the nasopharynx and overlying brain at the level of the foramen lacerum.
Figure 4 Schema showing the position of the levator muscle (1) with respect to the Eustachian tube, superior tubal ligament (5), and two components of the tensor veli palatini muscle—one that tenses the palate (2) and another that pulls the hook of the Eustachian cartilage downward (3). The superior pharyngeal constrictor muscle (4) is also shown.
Figure 3 Lateral view of the nasopharynx and oropharynx to show superior constrictor of the pharynx (6), which is shown attaching to the hamulus and pterygomandibular raphe and continuing on as the buccinator (7). The levator muscle (3) and the tubal cartilage (1) are shown passing through the sinus of Morgagni. The tensor muscle (2) is shown descending and passing around the hamulus (8). Other structures shown are: medial pterygoid muscle (5); salpingopharyngeal muscle (4); hard palate (10); middle constrictor muscle (9); lingual nerve (11); palatoglossus (12); and facial artery (13). Source: From Ref. 54.
of periosteum. The lateral walls of the nasopharynx are made up mainly of the structures in the foramen of Morgagni—that gap that extends between the clivus and the superior border of the superior constrictor. The Eustachian tube and its pharyngeal elaboration, the torus tubarius, along with the tensor and levator muscles of the palate, are the principal structures occupying this gap (Fig. 4). They are leashed together by the pharyngobasilar fascia. At the anterior aspect of the lateral extent of the nasopharyngeal roof, just medial to the region of the pterygoid plates, is the foramen lacerum. The inferior portion of the foramen is obliterated by a cartilage plug, but the superior part harbors the internal carotid artery as it sweeps out of the petrous apex into the cavernous sinus. The proximity of the internal carotid artery and cavernous sinus, and the presence of the brain stem on the intracranial side of the clivus had until recently precluded the use of surgery for lesions in the nasopharynx. Over the last two decades, surgical procedures on the clivus such as those developed by Fang and Ong (9) and Crockard (10) and Donald and Bernstein (11) have been developed to resect nasopharyngeal neoplasms. The transpalatal approach by Fee (12,13), which gives limited but adequate access for small tumors, has been complimented by the procedures of Cocke and Robertson (14), Wei (15,16) and Janecka (17) who approach the nasopharynx from the mid-face and those of Sekhar and Schramm (18), Fisch and Yasergil (19), who gain the exposure of this region from the lateral aspect. This latter approach is the topic of this chapter. These procedures have enabled surgeons to achieve
308
Donald
Figure 6 Varieties of sphenoid sinus pneumatization. Source: From Ref. 55.
Cavernous Sinus and Clivus
Figure 5 The pituitary gland, its immediate relations, and physiologic compartments.
complete resection of even squamous cell carcinoma at this site.
Sphenoid Sinus The sphenoid sinus resides in a unique area in the head, occupying its anatomic center. Lying above it, suspended from the region of the hypothalamus in the sella turcica, is the grand orchestrator of bodily hormonal function, the pituitary gland (Fig. 5). It is related laterally to the cavernous sinus and its contents. Posterior to the sphenoid sinus is the posterior extent of the basisphenoid and its articulation with the basiocciput; these latter two bones are collectively called the clivus. The anterior surface of the sinus contains its ostium, emptying into the sphenoethmoidal recess in the posterior vault of the nasal chamber. The sphenopalatine artery runs under the sinus ostium and may be a source of troublesome bleeding if not adequately coagulated prior to severance. In cases of marked pneumatization, the thick floor of the sinus may form a small part of the anterior roof of the nasopharynx. The inferior aspect of the anterior sinus wall forms the sphenoid rostrum, which is at the posterior most extremity of the roof of the nasal cavity. The sphenoid sinus may be highly variable in its extent (Fig. 6). It is described in relation to the sella turcica, which is a vital consideration in the transsphenoidal approach to the pituitary. Pneumatization may consist merely of a dimple on the anterior surface of the basiocciput, so called conchal pneumatization. In the highly pneumatized sphenoid, the air cell can extend into the base of the pterygoid plate and even into the lesser wing of the sphenoid bone. Presellar and especially postsellar pneumatization will permit a satisfactory approach to the gland. Although the optic nerve is usually related to the sphenoid sinus high on its lateral wall, in certain individuals the sinus actually wraps around the optic nerve (Fig. 7). This is especially true if the patient has a large sphenoethmoidal cell, the “cell of Onodi.” Before extensive surgery in this area, a careful study of the CAT scan of the sinuses and skull base in the coronal view is mandatory to establish the nature of this anatomic relationship.
The cavernous sinus lies directly against the bone of the lateral wall of the sphenoid sinus (Fig. 8). The dura that forms each sinus is derived from the sphenoidal periosteum. Each cavernous sinus extends from the orbital apex anteriorly to the petrous ridge posteriorly. The petrous apex terminates in the floor of the sinus. A plethora of veins drain into it, making it vulnerable to a variety of infection sources. The cavernous sinus connects to both the superior and inferior petrosal sinuses, as well as the sphenoparietal sinus. Further anteriorly, connections to the superior and inferior orbital veins establish a direct connection to the veins of the face. The neurovascular foramina in the floor of the cavernous sinus establish routes to the paranasal sinuses, the principal vein being the connection to the pterygoid plexus through the foramen of Vesalius.
Figure 7 Anatomic specimen cut in coronal section through the sphenoid sinus, illustrating the sinus surrounding about 200 degree of the left optic nerve. Posterior facing surface. (1) Superior compartment in the sphenoid sinus; arrow demonstrates foramen communication with the main body of the sinus. (2) Optic nerves separated by a thin plate of bone from the sphenoid sinus, which wraps around it on the left side for 280 degree. (3) Left sphenoid sinus larger than the right. (4) Right sphenoid sinus. (5) Pterygoid extensions of the sphenoid sinuses. (6) Greater wings of the sphenoid bone. (7) Cleft palate defect. Source: From Ref. 56.
Chapter 20: Infratemporal/Middle Fossa Tumors
Figure 8 Cavernous sinus showing connections across the midline via the circular sinus and the basilar plexus.
The foramen lacerum, the conduit for the internal carotid artery, is also in the sinus floor. In addition, there is a number of bridging veins that go from the cerebral surface to the cavernous sinus. The commonest ones are from the anterior and middle cerebral veins (Fig. 9). Each of the dural venous cavernous sinuses is connected to its fellow on the opposite side of the sphenoid sinus by the circular sinus. The circular sinus, although it encircles the pituitary stalk, is in reality a labyrinth of blood-filled channels and spaces rather than a single venous conduit (Fig. 8). The second connection across the midline is through the basilar plexus, which lies over the cranial surface of the clivus (Fig. 8). Running through the cavernous sinus are the optic nerve; the third, fourth, and sixth cranial nerves; the first, second, and occasionally the third branches of the trigeminal nerve; and the internal carotid artery (Fig. 10). The artery may indent the posterolateral wall of the sphenoid sinus and appear from the interior perspective of the sinus, as a bulge. The third cranial nerve (the oculomotor) lies superiorly in the superior aspect of the lateral dura of the cavernous sinus encased in a double layer. The fourth cranial nerve (trochlear nerve) is much smaller than the oculomotor and lies just below it. The clivus is composed of the basisphenoid bone anteriorly and the basiocciput posteriorly. The two bones are
Figure 9 Venous connections of the cavernous sinuses and the bridging cerebral veins leading into it.
309
Figure 10 Cavernous sinus internal carotid artery and cranial nerves.
connected by a synostosis. Much of the basisphenoid is often occupied by the sphenoid sinus. The posterior wall can be perilously thin providing a slight barrier between the sinus and the brain stem.
Undersurface of the Temporal Bone To most otorhinolaryngologists, the anatomy of the temporal bone is well understood. However, unless they are engaged in considerable inner ear surgery, most practitioners find that their comprehension of the inferior surface of the bone (Fig. 11), thoroughly studied during residency training, has dimmed with time. Although the temporal bone itself is shaped roughly like a triangle, with the base situated laterally and the apex at the petrous tip medially, it may be helpful to envision its undersurface as a rectangular area canted at roughly 45 degrees to the side of the skull (Fig. 12). The lateral corners are bounded by the articular eminence anteriorly and the anterior lip of the occipital condyle posteriorly. Although the inferior surface of the temporal bone makes 90% of this area of the skull base, the sphenoid anteromedially and the occiput posteromedially make small contributions.
Figure 11 The inferior surface of the temporal bone. Source: From Ref. 57.
310
Donald
(A)
Figure 12 (A) The undersurface of the skull base. This area is roughly rectangular and is canted anteromedially at 45 degree. Abbreviations: LPP, lateral pterygoid plate; AE, articular eminence; FL, foramen lacerum; FO, foramen ovale; PT, petrous tip; GF, glenoid fossa; JF, jugular foramen; OC, occipital condyle; MT, mastoid tip. (B) Anteroinferior view of temporal bones. Source: B from Ref. 58.
Chapter 20: Infratemporal/Middle Fossa Tumors
Just medially to the mastoid tip, the posterior belly of the digastric muscle takes its origin and receives its innervation directly from the main trunk of the facial nerve. A notch medial to the muscle houses the occipital artery. This artery can be responsible for considerable hemorrhage during surgery if not well controlled. As further progress is made anteromedially, the styloid process is seen with the origin of the styloid muscles (the so called muscles of Riolan comprised of the stylohyoid, the styloglossus, and the stylopharyngeus) and stylomandibular ligament. Just deep and posterior to it are the stylomastoid foramen and the emergence of the facial nerve. Directly medial to the styloid process is the jugular foramen, and just deep and anterior to it, separated by a thin ridge of bone, the carotid canal. Posteromedial to the carotid is the occipital condyle and its articulation with C1. Posteromedial to the carotid canal and under the lip of the occipital condyle is the foramen of the hypoglossal canal, the iter of the hypoglossal nerve. The vagus, glossopharyngeal, and spinal accessory nerves exit the skull adjacent to the medial surface of the jugular vein, through the incisura between the temporal and occipital bones. Returning more laterally to the anterolateral corner of the skull base rectangle, and proceeding from the articular eminence medially, the remaining anatomy of this complex region is revealed. Just posterior to the eminence is the condyle of the mandible in the glenoid fossa. The glenoid fossa lies directly underneath the lateral aspect of the middle cranial fossa. Its medial extent lays anteromedial to the foramina for the great vessels. A fissure cut coronally, directly through the center of the glenoid fossa plate, will expose the Eustachian tube. Surgical removal of the condylar head is essential to the complete exposure of the middle fossa floor. With the condylar head removed and sighting from lateral to medial, the undersurface of the temporal bone appears to be a steep triangle. Anteriorly, the lateral pterygoid muscular insertions track into the infratemporal fossa, and these, as well as the temporal muscle, form the anterior wall. The posterior wall is formed by the tympanic bone and its vaginal process. At the blunted apex of the triangle, the beginning of the cartilaginous Eustachian tube, the tensor palatini anteriorly and the levator palatini posteriorly are located. At the anterior most aspect of this apex is the sphenoid spine. Adjacent to the medial surface of the sphenoid spine is the foramen spinosum, which contains the middle meningeal artery. The sphenoid spine is also one of the points of origin for the tensor muscle of the soft palate, which courses along and takes further origin on the lateral plate of the cartilaginous portion of the Eustachian tube on its way to the nasopharynx.
The Infratemporal Fossa and Pterygomaxillary Space The infratemporal fossa is a space on the side of the head that begins superiorly at the temporal line at the inferior limit of the temporal fossa and lays deep to the zygomatic arch just in front of the “skull base rectangle.” The temporal fossa houses the temporalis muscles, while the pterygoid muscles and the internal maxillary artery (Fig. 13) reside in the infratemporal fossa. The medial wall of the fossa is the squamosal portion of the skull superiorly and the lateral maxillary sinus wall and lateral pterygoid plate inferiorly. The fossa leads into the pterygomaxillary space through the fissure of the same name in the middle of this medial wall. The pterygomaxillary space (Fig. 14) is directly behind the maxillary sinus, just under the medial aspect of the skull base and just above the pterygoid plates. The internal maxil-
311
lary artery enters the space following its passage between the two heads of the internal pterygoid muscle. Upon reaching the space, it ramifies into its terminal branches. The maxillary nerve, gaining entrance via the foramen rotundum, traverses the space while arborizing into branches that pass together with their accompanying vessels through the various exiting foramina. The vidian canal enters the pterygomaxillary space posteriorly, carrying the amalgamation of the deep petrosal and greater superficial petrosal nerves (Fig. 15). Although tumors rarely take origin in this space, neoplasms originating in the nasopharynx, buccal space, maxillary sinus, and infratemporal fossa often find their way to this site. The foramen ovale with its contained nerve, the mandibular branch of the trigeminal, is situated slightly anterior and medial to the middle meningeal artery. The medial most limit of the foramen approximates the beginning of the lateral pterygoid plate. Now the anteromedial apex of the skull base rectangle has been reached. An imaginary line subtended from the medial pterygoid plate to the anterior limit of the occipital condyle marks the lateral wall of the nasopharynx.
PATHOLOGY Although this chapter is intended to deal mainly with malignancies, some benign tumors—because of their location and biologic behavior—are considered malignant by position and proclivity. This is particularly true of tumors located at the skull base. For instance, juvenile nasopharyngeal angiofibroma—a vasoformative tumor of vascular channels and spaces and intervening fibrous connective tissue (Fig. 16)—is more prevalent among prepubertal males, often is hormonally dependent, and can remain occult for some time until presenting with nasal obstruction or epistaxis (Figs. 17 and 18). The tumor arises in the sphenoid bone near the origin of the pterygoid plates (20) presumably from totipotential cells that differentiate into the characteristic fibroblasts and endothelially lined vascular spaces of this tumor. It slowly expands into the nasopharynx and then extends into the paranasal sinuses, infratemporal fossa, skull base, and even the anterior and middle cranial fossae. The bony erosion produced by the tumor is probably a pressure phenomenon, and large protrusions often occur through the osseous barriers in its path. The characteristic presenting symptom of profuse epistaxis may be absent or of reduced intensity, masquerading as a routine “nuisance type” nosebleed. Such tumors may assume considerable size, as in the patient shown in Figure 18, who presented with total nasal obstruction, a mass in the cheek and palate, and proptosis. As intracranial involvement occurs, branches of the internal carotid artery make contributions to an already abundant blood supply. Chordoma, a locally malignant but rarely metastasizing tumor of notochordal origin, is one of the tumors that can be best managed by partial clivectomy. It arises from cells that are remnants of the embryologic notochord. The nasopharyngeal variety accounts for about one-third of these tumors, which take origin in the clivus at the suture line between the basisphenoid and the basiocciput. According to Binkhorst et al. (21), chordomas can also arise from the dorsum sellae, the retropharyngeal area, the apical ligament of the odontoid, and the nucleus pulposus of the cervical vertebrae. Because of their slow growth, they often do not present until they have compromised cranial nerves either in the optic or otic systems (22,23). Histologically, they are composed of trabeculations of polygonal cells containing eosinophilic
312
Donald
Figure 13 The infratemporal fossa. (A) Bony landmarks. (B) The infratemporal fossa contents. Source: A from Ref. 59. B from Ref. 60.
Chapter 20: Infratemporal/Middle Fossa Tumors
Figure 14 The pterygomaxillary space. Source: From Ref. 61.
granular or homogeneous cytoplasm. Cytoplasmic vacuolations are present in foci throughout the tumor and are thought to represent cellular aging (23). This pathognomonic vacuolated cell, the so-called physaliphorous cell, is characteristic of chordoma. These grossly lobulated tumors have a tendency to erode bone and invade soft tissue structures. Envelopment of cranial nerves and the internal carotid artery, together with aggressive invasion of the dura, renders complete resection of these lesions extremely difficult. Until recently, total excision was considered impossible, and palliative surgery and radiotherapy were often the only choices of therapy. Refinements in craniofacial surgery and the advent of transclival resection and the far lateral approach have made possible total resection in heretofore hopeless cases. Connective tissue tumors such as fibromas, schwannomas, lipomas, and their malignant counterparts can inhabit the skull base region. Both benign and malignant forms can erode bone and extend intracranially. Moreover, neither type can be differentiated on the basis of local behavior alone. Sarcomas of the nasopharynx or those invading the skull base from adjacent regions have a uniformly dismal prognosis because of local recurrence. On the other hand, the benign forms, if totally removed, have an excellent prognosis. Although the major problem with sarcomas is local recurrence, occasionally regional lymphatic, but especially distant bloodborne metastases pose a significant problem. The erosion of bone is not the only route of spread. The neural and vascular foramina of the middle fossa floor also provide portals of tumor intrusion. Often these tumors remain occult until compromise of the nerves traversing these foramina produces symptoms suggesting their pres-
313
ence. Schwannomas of the nerves of the jugular foramen, when high in the neck, may be mistaken for neoplasms in the tail of the parotid. An MRI and CAT scan of the upper neck will make the differentiation (24). Malignant connective tissue tumors are rare. Leiomyosarcoma, fibrosarcoma, liposarcoma, myxoma, malignant histiocytoma, rhabdomyosarcoma, and malignant schwannomas have all been reported and commonly occur in the recesses near the skull base (Fig. 19). Malignancies of blood vessel origin, such as hemangiopericytoma and malignant glomus jugulare tumor, are also seen. Hemangiopericytomas, originating in the pericytes of Zimmerman, have a dismal prognosis when intracranial spread occurs, since local invasion to this extent usually means that there are concomitant distant metastases. The orbit and nasopharynx are the most common sites of origin in the head and neck (25). The skin and muscle account for the majority of tissues primarily involved, and about 25% of all such lesions occur in the head and neck. Hemangiopericytoma is a rapidly growing, locally aggressive tumor with a high local recurrence rate. Regional lymph node metastases are uncommon. The histologic diagnosis is sometimes difficult. The tumor cells are located between the flattened endothelial cells of the numerous capillaries of the tumor and the investing reticulin layer (26). Reticulin stains are often necessary to clarify this relationship. Curiously, there is some electron microscopic evidence indicating a similar origin of hemangiopericytoma and subungual glomus tumors (27). Malignancy is difficult to determine histologically. Mitotic figures and anaplasia are features common to both benign and malignant varieties and are inaccurate predictors of biologic activity. Disturbance of the relationship between the tumor cells to the reticulin layer has been reputed to be a sign of malignancy (28). Leiomyosarcoma is very uncommon and is rarely seen in the head and neck. Most of these tumors arise in the subcutis, probably from the erector pili muscles, but occasionally lesions will develop from the wall of the internal jugular vein. Rhabdomyosarcoma is most commonly seen in the orbit in children, with the next most frequent site being the nasopharynx. It is not the purpose of this chapter to present an exhaustive treatise on soft tissue tumors, since this confusing group of neoplasms is expertly dealt with by a number of other recognized authorities (29–31). However, some attention should be given to the fibrous neoplasms, as they are a confusing m´elange of tumors and proliferations. Not all types are common problems in the region of the skull base, but an understanding of the natural history of these tumors as a group is necessary to have a clear picture of them as a pathologic entity. In describing tumors of connective tissue cell origin, Conley eloquently states, “Although these specific tissue entities lend themselves to an arbitrary academic classification, their dynamism for capricious primitive cellular development, functionalism, and partial mutation seems to create critical obstacles to their accurate microscopic classification” (32). Batsakis (29) roughly classifies fibrous lesions as benign, malignant, or indeterminate. Of the last group he explains, “Classifications of this group of tumors (proliferations and neoplasms) have often been rambling monstrosities with terms such as fasciitis, nodular fasciitis, infiltrative fasciitis, pseudosarcomatous fibromatosis, aggressive fibromatosis, cellular keloid, differentiated or Grade 1 fibrosarcoma, and nonmetastasizing fibrosarcoma, either lumped together or sharply defined by ill-defined and often intangible criteria.”
314
Donald
Figure 15 Pterygopalatine fossa with posterior foramina and surrounding area in frontal section seen from front. [1, superior orbital fissure; 2, oblique septum of the sphenoidal sinus; 3, left sphenoidal sinus; 4, left optic canal; 5, floor of the middle cranial fossa; 6, sphenoidal lamina; 7, lateral plate of the pterygoid process; 8, dorsum sella and foramen rotundum: millimeter strip; 9, optic canal and anterior aperture of the pterygoid canal; 10, palatovaginal canal and palatine (sphenoidal process).]
The differentiation of fibroma from fibrosarcoma on a histologic basis is often extremely difficult, since the usual criteria of mitotic activity, cellular pleomorphism, nuclear hyperchromatism, and extracapsular invasion cannot be reliably used to make this distinction. The presence of metastases, of course, definitively designates the lesion as malignant, but this is an uncommon occurrence in fibrosarcoma. Very often the differentiation has to be made on the basis of biologic activity. The pathologic diagnosis then emerges from a composite of clues derived from historic, radiographic, and histologic data. Very often the surgeon who is managing the case
Figure 16 Photomicrograph of juvenile angiofibroma showing numerous vascular spaces in a fibrous tissue stroma (H & E, ×220). Source: From Ref. 62.
must make the definitive diagnosis, since this surgeon alone is cognizant of the aggressiveness of the specific tumor under treatment. Figure 20 portrays even more graphically the confusion that exists over the classification of fibrous tumors, although it
Figure 17 Coronal MRI of a 17-year-old boy with a large nasopharyngeal angiofibroma beginning near the superior aspect of the sphenoid bone near the origin of the pterygoid plates. Note tumor extending into the infratemporal fossa and the middle cranial fossa.
Chapter 20: Infratemporal/Middle Fossa Tumors
Figure 18 14-year-old boy presenting with epistaxis, mild proptosis, cheek swelling, and nasal obstruction.
does in some way provide a framework to work by. Fibromas are difficult to diagnose and are quite uncommon. Many of the tumors formerly bearing this appellation were in fact histiocytomas. Most of these tumors occur in skin and are rarely seen in the region of the skull base. An illustrative case is a patient who had a desmoid tumor of the infratemporal fossa that invaded the middle cranial fossa floor and extended into the middle fossa and cavernous sinus. This patient presented with impairment of all branches of the trigeminal nerve on the affected side and a frozen globe. Despite two aggressive attempts at resection including the entire cavernous sinus, the temporal lobe dura, dissection of the internal carotid artery, removal of the sphenoid sinus and the clivus, resection of the infratemporal fossa, nasopharynx and oropharynx, negative margins could not be obtained. Postoperative irradiation was given with little effect on the residual tumor. The patient was a prisoner and extended follow-up was impossible. This case illustrates the problem of classification of these fibrous neoplasms whose behavior is clearly malignant. Fibrosarcoma was formerly one of the most commonly diagnosed soft tissue malignancies, often in error. An illustrative case is that of a patient who had the appearance of a
Figure 19 CT scan showing a leiomyosarcoma invading the third branch of the trigeminal nerve and extending through the foramen ovale.
315
slowly growing mass in the left neck that suddenly, over a period of 3 days, became rapidly larger. The patient had severe pain over the mass, paralysis of the left side of the tongue, and a complete left vocal cord paralysis. Surgical exploration revealed a fish-flesh appearing, hard mass encasing the common and internal carotid arteries and all the lower cranial nerves in the neck. Diagnosis was made variously as an “aggressive fibrotic process consistent with nodular fasciitis” and fibrosarcoma by a number of consultants from different institutions. The patient had a rapid and complete resolution of the mass with a course of steroid medication. At eight years follow-up, there was no mass and the cranial palsies had resolved. In recent times, fibrosarcomas have been determined to make up approximately 0.5% to 5.5% of all soft tissue malignancies (33). The degree of histologic differentiation usually determines the biologic activity. The well-differentiated tumor rarely metastasizes and tends to be less locally aggressive. Regional and distant metastases, as well as aggressive local spread, characterize the poorly differentiated form. The well-differentiated variety demonstrates an admixture of fibers and cells. The cells are fairly uniform, mitoses are infrequent, and nuclear hyperchromatism is occasionally seen. Varying degrees of cellular anaplasia designate the malignancy of the undifferentiated forms. The metastatic rate in the latter is around 20% to 25% (18,19) and is usually hematogenous. A much more common tumor in the skull base area is the schwannoma. Electron-microscopic and histochemical evidence definitely points to the Schwann cell as the origin of both schwannomas (also called neurilemomas) and neurofibromas (34). Clinically and grossly, the two have characteristic differences. One of the most important, from the surgical standpoint, is that the neurofibroma has neural elements that traverse the tumor, whereas the schwannoma, although attached to a nerve, has no actual nerve fibers traversing it. Neurofibromas are characteristic of von Recklinghausen’s disease and carry a significant potential for malignant degeneration. Histologically, the Antoni A and B configuration and the presence of Verocay bodies are characteristic (Fig. 21). Although the configuration of a central cytoplasmic mass encircled by palisaded nuclei that together comprise the Verocay body is occasionally seen in neurofibromas, the usual picture is of an unencapsulated tumor composed of fibers and spindle cells insinuated within the nerve substance. The fibrous component is commonly randomly distributed, and the cellular nuclei take on a serpiginous configuration. The Schwannomas and some paragangliomas such as carotid body tumors and glomus vagale tumors may take origin in the poststyloid part of the parapharyngeal space and spread superiorly to the infratemporal fossa. Indeed some anatomists, such as Ciszek34A , feel that the infratemporal fossa is simply the superior extension of the parapharyngeal space. Batsakis (29) prefers to designate malignancies of nerve cell sheath origin as neurogenous sarcomas. Malignant schwannoma would appear to be a misnomer, as clear evidence of malignant degeneration in a schwannoma is wanting. The appearance of some cellular pleomorphism, an occasional mitotic figure or multinucleated cells, must be recognized as a regressive step and not as a sign of malignancy. Neurofibromas appear to be the principal source of malignancy among neural tumors of the head and neck. Histologically, the differentiation of neurogenous sarcoma from fibrosarcoma is almost impossible. The origin of the tumor in a nerve is the prime distinguishing feature. These malignancies
316
Donald
THE FIBROBLAST AND THE HISTOCYTIC FACULTATIVE FIBROBLAST (HFF)
“CLINICALLY” MALIGNANT
“CLINICALLY” BENIGN TUMORS SCAR-LIKE PROLIFERATION Proliferating Scar
Keloid
Atypical fibroxanthoma of skin FASCIITIS
Certain forms of congenital torticollis
Dermatofibroma
Nodular (pseudosarcomatous Proliferative myositis Parosteal fascutis (?) Myositis ossificans
Differentiated fibrosarcoma
“CLINICALLY” AGRESSIVE AND OFTEN INDETERMINATE TUMORS The musculo-aponeurotic fibromatoses EXTRA-ABDOMINAL DESMOID
SUBMUCOSAL PROLIFERATIONS Oral submucous fibrosis
Histiocytoma
Malignant fibrous xanthoma (fibroxanthosarcoma)
NON-DESMOIDAL FIBROMATOSIS Aggressive
Inflammatory” ‘‘polyps and epulis”
Gingival hypertrophy
Undifferentiated fibrosarcoma
“Juvenile” FIBROUS XANTHOMA
(?) Pseudosarcomatous reactions associated with squamous cell carcinomas
Multiple congenital fibromatosis Certain forms of congenital torticollis ? Fibrous hamartoma of infancy
“Dermatofibrosarcoma” protoberans
Mixed fibrous xanthomas
Figure 20 Schematic attempts to classify fibrous tumors and fibroproliferative diseases of the head and neck. Source: From Ref. 63.
frequently metastasize and have rapid local spread. Only wide radical local excision provides any hope of salvage. Sarcomas are rare in the nasopharynx, and squamous cell carcinoma is by far the most common malignancy seen, usually of a poorly differentiated cellular type or as a lymphoepithelioma (WHO III). For some time, controversy over the precise categorization of lymphoepithelioma and transitional cell cancer raged. The electron-microscopic findings of keratin granules and desmosomes (35,36) have definitively classified these tumors as variants of epidermoid carcinoma. Their biologic behavior is similar to that of poorly differentiated squamous cell cancers, and, like them, they are more radiosensitive than their more differentiated relatives. Both lymphoepithelioma and transitional cell cancer may be grossly exophytic, ulcerative, or infiltrative. The occult infiltrative type is the most common (Figs. 9–22). Following the palatine tonsil the nasopharynx is the most common site of lymphoma. It is frequently accompanied by cervical lymphadenopathy.
Figure 21 Photomicrograph of a schwannoma (×130).
Nasopharyngeal malignancies in general are uncommon, making up only 0.25% or so of all cancers. In the population at large, its overall incidence is only about 0.0005% (37). A notable exception is the high incidence of nasopharyngeal cancer in China, where it is responsible for about 15% of all deaths due to malignant disease (35). This is not true for Chinese born in other countries. Although this incidence of nasopharyngeal malignancy is still high, it diminishes among each successive generation of Chinese born outside their native country. The role of the Epstein–Barr virus in the etiology of this disease is of considerable interest. A common association between elevated titers of Epstein–Barr antibody commonly coexists with the disease. Further, Coates et al. showed a decrease in antibody titers in those cured of their disease (38). An association between nasopharyngeal carcinoma and the consumption of salt fish has been made (39). This is especially strong when this consumption begins in early childhood.
Figure 22 Autopsy specimen showing massive intracranial extent of an infiltrating nasopharyngeal carcinoma arising in the fossa of Rosenm¨uller.
Chapter 20: Infratemporal/Middle Fossa Tumors
The most common presenting sign in nasopharyngeal carcinoma is a lump in the neck. Conversely, one of the most common of the unknown primaries in patients who present initially with a lump in the neck is nasopharyngeal carcinoma (40). Fletcher and Million (41) found only 16 cases out of a total of 112 (14.3%) that had no metastasis to the neck. Vilar (42) reported 21/24 (87.5%) patients with nodal involvement, and another authority (43) quotes 90% of patients with unilateral and 50% with bilateral disease. The lymphatic drainage system of the nasopharynx offers a number of different routes of metastatic spread. There are from one to three high laterally placed retropharyngeal lymph nodes that are the principal filtering stations of the lymphatic channels draining the nasopharynx. The lymph proceeds from these to the upper internal jugular nodal group or to the nodes of the posterior cervical triangle. In some instances, the high nodes are bypassed and the first manifestation of nasopharyngeal tumor is an enlarged node in the posterior triangle or, less commonly, in the upper internal jugular chain (44). The original study done by Rouvi`ere (45) over 40 years ago clearly delineating the lymphatic system in the head and neck has not been improved upon since. Squamous, adenoid cystic, and adeno carcinoma; malignancies of the skin such as squamous cell and basal cell carcinomas as well as melanoma; those of the paranasal sinuses such as adenoid cystic carcinoma and adenocarcinoma; and those of the parotid gland such as adenocarcinoma and acinic cell carcinoma will be discussed in more detail in the section on pathophysiology.
Pathophysiology of Spread of Skull Base Tumors The routes of spread of malignant neoplasms through the skull base into the intracranial cavity follow a variety of routes. Although direct invasion by bone or cartilage erosion is a common method of spread, direct extension through the natural avenues provided by neural and vascular foramina provide easy access to the subdural spaces. Direct spread by bony erosion is commonly encountered in paranasal sinus malignancies as the bone of the fovea ethmoidalis and cribriform plate area is quite thin. Similarly, tumor spread through the lateral walls of the sphenoidal sinus to the dura of the cavernous sinus is facilitated by the thinness of the bone while extension posteriorly through the clivus is uncommon because of its thickness except in cases of extensive postsellar pneumatization. Perineural spread of tumor through the neural foramina of the skull base is probably the commonest method of spread. Adenoid cystic carcinoma has this property as its pathognomonic characteristic. Most malignant tumors of the head and neck can on occasion spread along nerve sheaths, but squamous cell and basal cell carcinoma and malignant melanoma are most likely to acquire this propensity. These latter tumors when affecting the skin may spread along the branches of the trigeminal nerve through their respective foramina and into the Gasserian ganglion. Squamous cell cancers of the tonsil, inferior alveolar ridge, and even the lip and tongue may spread along the sheath of the inferior alveolar nerve to the mandibular division of the trigeminal nerve and through the foramen ovale to the ganglion. Tracking of tumor along the optic and olfactory nerves provide access to the anterior cranial fossa. A curious property of this method of spread in the subcranial route is the phenomenon of skip areas. There may be a gap of 1 cm or more between tumor deposits along the nerve until it gains its intracranial course. Once the nerve has traversed the foramen, the tumor
317
Figure 23 Dural invasion by squamous cell carcinoma (H & E, ×40).
continues mainly in continuity almost entirely devoid of skip areas. The spread of tumor along vascular structures through the skull base foramina, in contrast to perineural invasion, is in continuity. The vessels most commonly involved are the internal jugular vein and the internal carotid artery. Spread along the basilar artery, the mastoid emissary vein, and the pterygoid venous plexus are examples of this avenue of spread. Dural invasion occurs following intracranial spread via the neurovascular foramina and also when tumor spreads by direct extension. It is curious that in some cases, tumor that has eroded even a considerable amount of bone may abut against but not directly invade the dura. Dural invasion is not a contraindication to successful surgical tumor exenteration. In our series, after surgical excision, dural involvement made no statistical difference in the 5-year tumor-free survival and local tumor control rates (46). Dura has the ability to contain malignant tumors for an extended period of time before they will invade brain substance. Furthermore, the tumor in dura tends to be compact and has a tendency to expand into the tissue by a pushing action (Fig. 23). This enables adequate control of tumor when only a surgical margin of only a few mm is taken. Brain invasion is uncommon even when there is extensive involvement of dura. Unlike primary malignancies of the central nervous system, which have multiple satellites of the tumor at a distance from the primary, cancers of the upper digestive tract tend to spread by direct tumor extension and tend to remain in continuity (Fig. 24). They behave similarly to their spread in dura in this regard. Usually, the tumor is preceded by a halo of necrosis and edema, which is apparent on the MRI (Fig. 25). After resection for brain invasion, survival rates differ little from those with dural invasion (46). The phenomenon of meningeal carcinomatosis (47,48), thought to be the result of spread of tumor cells in the CSF, is very uncommon in our experience having occurred in only three patients of our. Internal carotid artery involvement occurs by direct extension either from the neck, through the temporal bone, via the foramen lacerum or vidian canal, and finally in the cavernous sinus either from tumor breaking through the ethmoidal or sphenoidal sinuses or from the various venous foramina entering the cavernous sinus. Metastatic carcinoma in an upper cervical lymph node from an upper aerodigestive
318
Donald
Figure 24 Histologic section demonstrating area of edema and necrosis (arrows) preceding advancing tumor front (H & E, ×100).
tract primary is the usual source of tumor spread in the neck. Direct extension from a parotid gland tumor or from a deeply invasive tumor from the oropharynx or nasopharynx occasionally occurs. The cervical carotid is covered by a layer of adventitia that provides a barrier to tumor spread. The fibrous ring (Fig. 26) that encircles the carotid as it enters the undersurface of the temporal bone provides significant resistance to tumor penetration. In the carotid canal that traverses the temporal bone, the carotid is protected first by a layer of stout periosteum and then a layer of loose areolar tissue that carries some of the blood supply to the wall of the artery. Finally, a layer of adventitia covers the vessel. The media of the vessel is slightly thinner than the artery in the neck (Fig. 27). The periosteum and even the loose connective tissue and adventitia provide a good barrier to tumor penetration, especially by low-grade tumors. The cavernous carotid has a
Figure 26 Fibrous ring of the internal carotid artery at the carotid foramen.
Figure 25 Lateral magnetic resonance image showing area of edema and necrosis preceding tumor front (arrows) of poorly differentiated carcinoma.
Figure 27 Diagram showing histologic details of the internal carotid artery in the carotid canal of the temporal bone.
thinner media and very little surrounding connective tissue to prevent tumor invasion of the artery. The cavernous sinuses along with the petrous and cavernous carotid are highly controversial areas in terms of feasibility of resection. Its multiple venous connections and its intrinsic highly vascular structure make resection of the cavernous sinus a formidable surgical challenge. Only a limited number of neurosurgeons are equipped by training or experience to tackle this challenging anatomic area. Tumors invading the sinus tend to displace the vascular channels resulting in scant bleeding at the beginning of the resection and only when the margins of tumor are reached does hemorrhage become a problem. Tumors that extend across the midline either through the circular sinus or the basilar plexus are considered by our team to be inoperable.
Chapter 20: Infratemporal/Middle Fossa Tumors
SURGICAL PROCEDURES Preoperative Planning One of the most important steps in the performance of procedures in the infratemporal fossa and skull base areas is preoperative planning. A thorough history and physical examination including a comprehensive review of all past records is essential. The previous biopsies or histologic sections of any past surgery must be reviewed by the skull base team pathologist. All previous radiographs are reviewed and usually it is necessary to get more current films. MRI and CAT scans are obtained with a focus on the skull base. Axial and coronal views usually give the most useful information. Gadolinium contrast and, often, fat suppression software are needed with the MRI to provide optimal information. The marriage of PET and CAT scan information as the PET–CT scan provides interesting data but lacks precision in delineating the exact extent of tumor. Newer refinements will make this modality extremely useful. This technology is especially helpful in the prediction of recurrence as prior surgery and radiation therapy obscure the tissue planes as displayed in standard MRI and CAT scans making detection of recurrent disease difficult to impossible. If the tumor is close to the internal carotid artery, then a carotid arteriogram with a balloon test occlusion and a SPECT scan is done to determine the ability of the contralateral arterial supply to supply the brain on the test side if the carotid is sacrificed (49). If the carotid artery is sacrificed, the artery is replaced with a graft preferably the saphenous vein. Because of the exigencies of each case, there will be considerable variation in the approach and extent of the surgery. Once a comprehensive examination and investigation have been conducted, by both the head and neck surgeon and neurosurgeon, a presurgical conference should be scheduled. At this time, the skull base surgical team will establish the sequence of operative steps, including the incisions, the temporal order of intracranial and transfacial dissections, the method of tumor delivery, and finally the stages of reconstruction and closure. With careful planning, often the cranial and facial incisions can complement one another. Care must be taken not to compromise the cutaneous blood supply by poor flap design. In many instances, a single incision can afford a perfectly adequate exposure for both facets of the procedure. The use of curved incisions not only is more cosmetic but also enhances exposure. Determining the sequence of the steps of dissection often requires considerable deliberation. However, this should not take the form of a power struggle but should be a process of deciding how best the dissection should proceed. Reconstructive efforts have as a first priority a watertight dural closure and separation of the cranial cavity from the upper aerodigestive system. Dural grafting is often required. The experience of Ketcham et al. (50) of improved results with use of temporalis fascia over fascia lata as dural replacement is borne out by my own. Dural substitutes in the form of lyophilized dura and bovine pericardium can be used as well. Dural allografts have gained popularity, however, caution must be exercised in cases that have been irradiated or in whom prior infection had occurred, as a general principle. Whenever possible, the dural repair is covered with a flap of pericranium. This flap, the workhorse of reconstruction in the anterior skull base approaches, is usually not feasible in the middle fossa/infratemporal fossa operations. More commonly, the temporalis muscle with an attached cuff of pericranium is used. Considerable time is often required to ensure precise dural integrity. This is time well spent, as it is the key to
319
the prevention of cerebrospinal fluid leakage and subsequent meningitis. Isolation of the cranial cavity from the air and food stream may not be a problem. Sufficient soft tissue in the upper neck or around the skull base may exist to reinforce the dural repair. If not, the interposition of a flap may be necessary to seal off the defect. Neurosurgeons commonly employ alloplastic materials for re-establishing cranial continuity. However, these materials should only be used when there is good vascularization of the bed. The concomitant use of a free flap adds a good measure of insurance. Adequately vascularized tissue is essential for the take of the dural and skin grafts. A good general principle is for both surgeons—to be present during most of the operation. As the partnership matures, substantial assistance can be rendered between participants. Anticipation of problems can be better predicted and more efficient dispatch of the operative process can be affected. Sudden unexpected hemorrhage may be best handled by the complementary team member; when both are present, the success of this cooperative venture will be maximized.
Lateral Approach to the Skull Base The undersurface of the skull base, specifically the inferior surface of the temporal bone, which is in fact the floor of the middle cranial fossa, is best exposed through the lateral infratemporal fossa approach. Visualization is enhanced not only by mandibular condylectomy, but also by osteotomy and mobilization of the zygoma (51). The parotid gland is mobilized and the facial nerve not dissected unless the tumor takes origin in the gland or the gland is invaded by tumor extension. Retraction of the gland rarely results in facial nerve weakness and does not hamper the exposure. Fisch developed the zygomatic osteotomy approach (52), for enhancing the exposure of large acoustic neuromas and glomus jugulare tumors with extensive intracranial and upper cervical involvement. His primary motive was to eradicate extensive otologic disease. We have found this approach useful also for deep-seated lesions such as extensive deep lobe parotid tumors and extensive schwannomas originating from the nerves of the jugular foramen, as well as tumors that reach to or extend from the nasopharynx, sphenoid sinus, and even the pituitary fossa. Sekhar and Schramm (17) modified Fisch’s approach to enable the resection of the intracranial extension of subcranial neoplasms, especially those that are malignant. The approach that our team uses is very similar to that of Sekhar and Schramm with minor variations.
INCISION The incision is an adaptation of the modified Blair incision for parotidectomy [Fig. 28(A)]]. The incision extends from the vertex almost to the hyoid bone [Figs. 9–28(B)]. This is used for those tumors that are more anteriorly located such as deep lobe parotid tumors, carcinomas with perineural spread along the branches of the trigeminal nerve and recurrent nasopharyngeal carcinomas. A postauricular “C” shaped incision as described by Fisch (51) is used in those patients with more posteriorly located tumors such as clivus chordomas, temporal bone carcinomas, and glomus jugulare tumors with intracranial extension (Fig. 29). As the skin flap is dissected forward, the frontal branch of the facial nerve is preserved by incising and then elevating a semicircular flap of deep temporalis fascia (Fig. 30). The external auditory canal is transected and the auricle retracted posteriorly, if the latter is to be retained. A decision now needs to be made regarding preservation of the middle ear space. If the Eustachian tube is to be
320
Donald
Figure 28 (A) Standard preauricular infratemporal fossa–middle cranial fossa incision and flap elevation. Note the use of the Taney clips on the scalp for hemostasis.
transected in order to dissect the internal carotid artery, then eliminating the middle ear space and ablating the external auditory canal is seriously considered on the basis of patient age, the disease process, and whether or not the patient has been irradiated in the past. In young patients with juvenile nasopharyngeal angiofibromas invading the middle cranial fossa and the cavernous sinus, the middle ear is preserved and a ventilation tube inserted. In an older patient who has been previously irradiated, the likelihood of postoperative infection and the danger of exposing the middle fossa dura to the outside air and a chronically infected ear put the patient at constant risk of meningitis. The hazard is even greater if the internal carotid artery is exposed, especially if it has been grafted. In the majority of cases, the middle ear is ablated. When using the postauricular approach, ablating the middle ear space then the method of Fisch (51) is used (Fig. 31). As this effectively removes most of the middle ear
Figure 29 (A)Postauricular “C” shaped incision for access to more posteriorly located lesions. (B) Line drawing of posterior incision.
Figure 30 Incision of temporalis fascia to include the frontal branch of the facial nerve, precluding it from injury. Source: From Ref. 64.
mucosa and closes the external ear canal in two layers, the cervical extension of the incision is used not only for access to the parotid tail but also as a means of exposing the great vessels from the carotid bifurcation to the skull base. The internal jugular vein and the internal and external carotid arteries are identified and encircled with a vascular loop (Fig. 32). If a radical neck dissection is required, a lazy S extension is made at right angles to the cervical aspect of the incision inferiorly. As always, the junction is made behind the carotid sheath. If a radical neck dissection is required, it is done after a parotidectomy, when the latter is necessary. Often both the superficial and deep lobes are removed. The facial nerve is carefully dissected in the manner described by McCabe and Work (52). In resecting the deep lobe, the nerve is gently retracted with a Zimmer hook. During the rest of the dissection, the continuity of the nerve is preserved up to the point where further retraction puts it in danger of severe neurapraxia or avulsion. At this point, severance just beyond the point of main branching is done so that when continuity is restored (near the termination of the operation), subsequent reinnervation will result in differential function. Most usually the parotid gland does not need removal and the gland and the facial nerve are preserved intact, the gland retracted inferiorly and the nerve not dissected. The skin flap is elevated in the “face-lift” plane up to an imaginary line between the lateral orbital rim and the angle of the mandible. A radical neck dissection is done at this point, if required by the exigencies of the tumor. If not, the inferior aspect of the apex of the temporal bone is now approached in a stepwise manner from lateral to medial. The initial exposure is gained by excising the origin of the sternocleidomastoid muscle from the mastoid tip. A slicing motion with a scalpel blade or cautery against the
Chapter 20: Infratemporal/Middle Fossa Tumors
(A)
(B)
(C)
(E)
321
(D)
(F)
Figure 31 (A) Periosteal–fascial flap pedicled on cartilaginous canal. (B) External canal skin elevated off conchal cartilage and out to level of external auditory meatus. (C) Suture placed through canal skin and brought out of ear with a mosquito clamp. (D) External auditory canal skin closed. (E) Periosteal–fascial closure over external auditory canal. (F) Closure complete.
322
Donald
Figure 33 Diagram illustrating the cuts in zygomatic ostectomy. Bone is preserved for later reconstruction. Note holes drilled for wire ligation. At least two holes are required if plate osteosynthesis is used. Source: From Ref. 65. Figure 32 Vascular loops placed around the internal jugular vein and the internal and external carotid arteries.
mastoid cortex is often quicker than the standard dissection with the mastoid elevator. Deep to this muscle is the posterior belly of the digastric. This is elevated from its fossa with a Joseph elevator, with care taken anteriorly not to traumatize the facial nerve. The occipital artery in its groove is dissected free and ligated. It is wise to control this vessel with a ligature or suture ligature at the outset, since failure to do so leads to annoying hemorrhage. As one proceeds medioanteriorly, the styloid process comes into view. The stylohyoid, stylopharyngeus, and styloglossus muscles are removed and dissected anteriorly, with the facial nerve always kept in clear view. Care is taken while dissecting the stylopharyngeus muscle not to damage the glossopharyngeal nerve. The pharyngeal anesthesia and the even minor motor disturbance that may result from damage to cranial nerve IX may render the patient dysphagic, an especially troublesome consequence in elderly patients. Progress beyond the styloid becomes difficult because of the mandibular condyle and neck. The triangle gets tighter as the dissection proceeds more deeply. Resection of the mandibular condyle is delayed until after the next two steps in the operation.
TEMPORALIS MUSCLE DISSECTION The first phase of exposing the infratemporal fossa is the elevation and transposition of the temporalis muscle. A 2-cm wide cuff of pericranium is incised around the circumference of the temporalis muscle. The muscle is then dissected from the temporal fossa down to the level of the zygomatic arch. The object is to eventually dissect the muscle down to its insertion on the coronoid process of the mandible.
In order to completely free up the temporalis muscle, it is necessary to detach the zygomatic arch. The earlier dissection of the deep layer of the temporalis fascia makes exposure of the zygomatic arch much simpler and preserves the frontal branch of the facial nerve.
ZYGOMATIC OSTECTOMY The bony incisions in the arch are outlined as seen in Figure 33. The posterior incision is made just in front of the articular eminence of the temporal bone and the anterior incision through the lateral orbital rim and posterior extent of the malar eminence. The fine bone-cutting needle of the Midas
Table 2 Physiologic Factors and Drugs Affecting Cerebral Blood Flow Factors during anesthesia causing an increase in intracranial pressure Straining during induction Increased blood flow Increased intrathoracic or intra-abdominal pressure Agents Halothane Methoxyflurane Trichloroethylene Enflurane Ketamine Factors during anesthesia causing a decrease in intracranial pressure Hyperventilation Osmotic diuretics Nitroprusside Atropine Agents Innovar Nitrous oxide Barbiturates
Chapter 20: Infratemporal/Middle Fossa Tumors
323
struction of the middle fossa floor. Removal of the condyle leaves no significant postoperative functional deficit except a mild wandering of the jaw to one side on wide opening. Postoperative trismus is most often related to dissection of the pterygoid muscles, dissection of the temporal–mandibular joint and not removing the condyle or postoperative radiation therapy. After the condyle is removed, the remaining soft tissue in the glenoid fossa is dissected free and removed so that only the bare bone of the fossa remains.
FORAMEN OVALE EXPOSURE
Figure 34 Holes being drilled and plates applied, then removed for later osteosynthesis with the completion of the operation.
Rex drill is ideal for these cuts. Prior to the cuts, miniplates are bent into the appropriate shape and pre-registered with drill holes (Fig. 34). The initial cut through the zygomatic arch just anteriorly to the articular eminence is begun following circumferential dissection of the zygomatic periosteum. A narrow malleable retractor is placed under the arch to protect the underlying muscle and internal maxillary artery and the cut made with the B5 cutting needle of the Midas Rex. A subperiosteal dissection is then done exposing the lateral orbital rim from well above the zygomaticofrontal suture to the superolateral extent of the malar eminence. The orbital periosteum is dissected inside the lateral wall of orbit posteriorly for a depth of about a centimeter. The narrow malleable retractor is inserted into the anterior aspect of the lateral orbit to protect the periorbita and orbital fat from injury from the bone needle. The bone needle cuts through the lateral orbital rim between the previously made drill holes for a depth of about 5 mm to be just inside the orbital rim. The needle is now directed inferiorly and the cut is made to about the level of the anterior extent of the inferior orbital fissure. The needle is now inclined posteroinferiorly to cut across the malar eminence just to where the arch begins. Once the zygoma has been removed, it is preserved in a saline-soaked gauze sponge for later replacement.
MANDIBULAR CONDYLECTOMY To complete the exposure of the infratemporal fossa, the condyle of the mandible is removed (Fig. 35). A “T” shaped incision is made through the temporal–mandibular joint, the periosteum at the mandibular neck dissected and a malleable retractor placed under the deep surface of the neck to protect the underlying internal maxillary artery from being cut. The internal maxillary artery gives off the deep temporal arteries, which are now the lone blood supply to the temporalis muscle. Its viability is essential if it is to be used for recon-
With a Dever retractor against the temporalis insertion on the coronoid process and the leading edge of the anterior aspect of the ramus of the mandible and the other soft tissues in the infratemporal fossa, a subperiosteal dissection is carried medially under the middle cranial fossa floor to the foramen spinosum and the foramen ovale. The subperiosteal dissection is continued anteriorly to the foramen ovale until the root of the pterygoid processes is exposed. Any remaining soft tissue lying on the lateral aspect of the lateral orbital wall is removed. The middle meningeal artery is clipped or tied then divided. The dissection around the subcranial side of the foramen ovale may stir up bleeding from small extensions of the cavernous sinus. Any bleeding must be gently and carefully controlled with light bipolar cautery, thrombinR soaked Gelfoam , and hemostatic gauze. Bleeding from the superior aspect of the pterygoid plexus of veins is not uncommon and judicious use of the bipolar cautery is exercised so as to preserve the two nutrient arteries to the temporalis muscle. The craniotomy that is to now ensue is a combined effort by the head and neck surgeon and the neurosurgeon. The goal of the craniotomy is to produce an “L” shaped bone flap where the vertically oriented long limb of the “L” is made up of a smaller portion of the greater wing of the sphenoid and the greater part the squamous part of the temporal bone and the short limb is made up of a portion of the middle fossa floor (Fig. 36). The inferior bone flap incision extends from the middle ear cleft posteriorly to the base of the pterygoid plates anteriorly. The superior potion of the flap (the vertical limb of the “L”) is usually relatively small but is fashioned large enough to permit adequate exposure and resection of involved dura or brain.
INTRATYMPANIC EXPOSURE Utilizing the dissecting microscope, the head and neck surgeon creates an anterior tympanomeatal flap in the external auditory canal skin going from 10 o’clock to about 7 o’clock. If the middle ear is to be ablated, then the entire canal skin and tympanic membrane is removed. The flap is elevated down to the tympanic annulus and the annulus prized out of its sulcus. The protympanum, the hemicanal for the tensor tympani, and the cochleariform process are identified (Fig. 37). A small cutting bur is used to make a cut along the course of the external auditory canal at about 2 o’clock. The exposure of the protympanum and the construction of the superior canal wall cut are facilitated by making a saucer-shaped excision in the zygomatic root widening the exposure of the attic (the Trojanowski maneuver) (Fig. 38). The bony incision begins just outside the canal and is carried down to dura. The cut is then carried at this depth medially down the ear canal to the tympanic annulus. The annulus is cut and the bur carried
324
Donald
Figure 35 Condylectomy: Lateral aspect of the temporomandibular joint capsule is opened and connected to a vertical incision in the periosteum of the mandibular neck. The condylar neck is transected and the condyle removed. The meniscus and attached soft tissue are removed.
through a short distance in the attic, across the hemicanal for tensor tympani well in front of the cochleariform process in order to avoid injuring the facial nerve. The pretympanic part of the bony Eustachian tube is cut to only a depth of about 2 mm in order to not to damage the underlying internal carotid artery. Inferiorly, the bony external auditory canal is incised at about 7 o’clock full thickness into the glenoid fossa until the annulus is met (Fig. 39). The annulus is cut across and a 2-mm deep fissure cut in the floor of the hypotympanum and the inferior part of the protympanum in a similar fashion to the superior cut. The cut is kept shallow in the floor of the middle ear to avoid accidentally cutting the jugular bulb. The cut through the inferior aspect of the external auditory canal and tympanic annulus is now directed along the glenoid fossa bone anteromedially again at a shallow depth until the sphenoid spine is reached. The middle meningeal artery is identified and clipped or ligated if not yet done so. The bone cut is now extended deeply to the level of the middle fossa
dura and carried forward to the foramen ovale. The final part of the subcranial portion of the bony incision extends from the anterior lip of the foramen ovale anterolaterally through the inferior aspect of the greater sphenoid wing just above the origin of the pterygoid plates.
CRANIOTOMY The neurosurgeon now makes the vertical portion of the craniotomy flap. A bur hole is usually made in the so-called “key-hole” at the pterion. The cut with the craniotome is made in the greater wing of the sphenoid and the squamous part of the temporal bone large enough to afford the exposure necessary to resect any involved dura or temporal lobe and exposure needed to dissect the cavernous sinus if invaded by tumor. The calvarial cuts join up with those made posteriorly in the 12 o’clock position in the external auditory canal and anteriorly at the level of the base of the pterygoid plates. The
Chapter 20: Infratemporal/Middle Fossa Tumors
325
Figure 36 (A) L-shaped craniotomy. Diagram outlining cuts. (B) Craniotomy flap removed from patient with extensive amount of temporal lobe dura involvement.
craniotomy flap thus completed is still attached by a thin bridge of bone in the middle ear. A careful dissection of the dura from superior to inferior is done while a gentle prying motion is exerted against the vertical part of the bone. The craniotomy flap fractures in a greenstick manner through the middle ear and glenoid fossa. It is important to construct the bone flap in this manner because the fracture through the middle ear goes through the protympanum and the bony Eustachian tube in the oblique axis of the temporal bone. When the bone flap is removed, a bulge in the posterior wall of the bony Eustachian tube is seen. This corresponds to the position of the internal carotid artery. Occasionally, this thin bony wall is dehiscent and the internal carotid can actually be seen pulsating.
COMBINED MIDDLE FOSSA/INFRATEMPORAL FOSSA RESECTION The exposure of the middle cranial fossa and this removal of the middle fossa floor up to the level of the foramen ovale enormously expand the exposure of the contents of the infratemporal fossa. How the subsequent dissection proceeds is dependent on tumor extent. Limited tumor invasion up the foramen ovale or the foramen rotundum may simply require excision of the nerve trunk at the level of the Gasserian ganglion in Meckel cave with resection of the overlying dura. If tumor invades the ganglion, then this structure is removed. If tumor tracks along the trigeminal nerve trunk, the neurosurgeon can trace the nerve along its short course on the
326
Donald
Figure 37 Anterior tympanomeatal flap retracted, exposing the tympanic opening of the Eustachian tube. The superior and inferior cuts into the protympanum are outlined.
posterior floor of the middle fossa, over the petrous ridge, and along its traverse of the posterior fossa up to the level of the brain stem. Invasion of the brain stem constitutes inoperability and the resection is terminated at this point. We have seen tumor extending this far in only rare instances. The characteristic skip areas pathognomonic of most tumors with perineural spread seen in the extracranial course of the cranial nerves when invaded by these neoplasms are rarely experienced in their intracranial course. Usually a 2 to 3 mm length of healthy nerve will provide an adequate margin of resection. Furthermore, the invasion of the dura overlying the involved nerves is usually of limited extent. If tumor invades the Eustachian tube, the tube can be removed with the cutting bur into the nasopharynx. This resection necessitates the dissection of the internal carotid artery. The entry of the artery at the fibrous ring (Fig. 26) is exposed in the upper reaches of the neck. Bone surrounding the artery is drilled up to the point at which it turns to its horizontal course posteriorly to the bony Eustachian tube. A bidirectional dissection (Fig. 39) removing bone over the artery seen at the bulge in the posterior wall of the Eustachian tube is directed along the horizontal part of the artery then joining that to the vertical part working superiorly from the carotid foramen. A diamond bur is used for that part of the bony canal that is next to the vessel. The
Figure 38 Entire view and close-up view showing bulge in posterior Eustachian tube.
Figure 39 Carotid dissections coming from two directions.
artery can be exposed up to the foramen lacerum, the cartilage plug contained in the inferior part of the foramen removed and the artery dissected up into the cavernous sinus (Fig. 40). Cavernous sinus invasion of cancers originating in the facial skin and upper aerodigestive tract is, for many surgeons, a contraindication to surgical intervention. Unquestionably the prognosis of such patients is much poor than other patients having skull base surgery. However, in our series the 3-year and better survival rate after cavernous sinus resection for malignancy is better than 20%. Although this survival rate is only one-half to one-third that of the usual survival for skull base resection, I believe it is worth the effort. The cavernous sinus resection is obviously the province of the neurosurgeon. The head and neck surgeon usually does the carotid exposure in the temporal bone using the dissecting microscope. The neurosurgeon dissects the sinus very carefully securing hemostasis by proceeding slowly using the bipolar cautery, Gelfoam soaked in thrombin, hemostatic gauze, and pressure with cottonoid. Since the cavernous sinus is a low-pressure system, hemostasis can be readily achieved with pressure. The tumor tends to replace that part of the sinus that it invades and bleeding is minimal during the
Figure 40 Internal carotid artery dissected from the carotid foramen to the cavernous sinus.
Chapter 20: Infratemporal/Middle Fossa Tumors
327
Figure 41 Redesigned cavernous sinus resection.
early phases of tumor removal. More vigorous bleeding is not encountered until the tumor has been mostly resected and the displaced vascular channels of the sinus are now encountered. If the resection can be confined to the anteroinferior regions of the sinus, usually the third and fourth cranial nerves can be avoided as they are located superolaterally between a double layer of dura. Similarly the sixth nerve can be avoided, if the tumor extension is limited to this anterior position. Unfortunately, the sixth cranial nerve is highly susceptible to damage even with minor manipulation. The lateral
Figure 42 Internal carotid artery replaced by Gortex graft from petrous temporal bone at foramen lacerum to upper neck.
rectus palsy thus produced by minor injury to the nerve often resolves over time. Because of the initial disappointing results of cavernous sinus surgery for direct invasion of malignancy (5year tumor-free survival 16.7% and local control rates 50%), the operation was redesigned to make the resection wider with larger margins of uninvolved tissue (Fig. 41). In addition, patients with malignancy are now often staged with an exploration of the opposite sinus to ensure the tumor had not crossed from the sinus with tumor involvement via the circular sinus or basilar venous plexus to the opposite side, which is negative or equivocal on the MRI or CAT scan. In the past 5 years, the redesigned operation outlined in Figure 41 takes a wider margin of healthy tissue including the petrous apex, the entire lateral wall of the sphenoid sinus, and the lateral one-third of the clivus and adjacent body of the sphenoid bone. The superior and inferior petrous sinuses, a portion of the basilar plexus and the carotid artery, if it is involved, are resected. The internal carotid, if sacrificed, is always grafted even if the patient has passed balloon test occlusion and SPECT scan. The resection of the internal carotid artery is even more controversial than the operation on the cavernous sinus. Our only long-term survivors of internal carotid sacrifice have been since the revision of the cavernous sinus operation. Prior to the new procedure, we had no survivors at 5 years and only a few at 2 years postresection. On the other hand, when the tumor was dissected from the carotid canal in the petrous temporal bone and the malignancy had not invaded the media of the artery then 25% of patients survived 5 years. Because the carotid canal in the petrous temporal bone is lined by a sturdy periosteum and the artery has a surrounding area of areolar tissue and adventitia (Fig. 27) there is a
328
Donald
(A)
(B)
(C)
(D)
Figure 43 Return of bone flaps and temporalis muscle used to reconstruct the middle fossa floor. (A) Temporalis muscle flap. (B) The temporalis muscle is placed across the craniotomy defect below the middle fossa floor and into the nasopharynx. (C) Craniotomy flap restored and muscle flap turned in. (D) Coronal view showing suture of pericranial attachment of the temporal muscle to the basipharyngeal fascia. Source: From Ref. 66.
significant barrier to tumor penetration, especially in cases of low-grade malignancy and those that tend to have a pushing margin such as adenoid cystic and acinic cell carcinoma. The preferred graft material for the carotid is saphenous vein and c (Fig. 42). The grafts usually the second choice is Gore-Tex extend from the neck to the take-off of the ophthalmic branch of the distal internal carotid. Closure of the wound begins with a dural closure as watertight as possible. Commonly, dura has been resected and a graft of fascia lata, temporalis fascia, lyophilized cadaver dura, or bovine pericardium is used for repair. Direct suturing is augmented with tissue glue. The floor of the middle fossa is closed by transposing the temporalis muscle across the middle fossa floor and suturing the pericranial cuff to the residual mucosa and muscle of the nasopharynx. The cranial bone flap and the zygoma are replaced (Fig. 43). Occasionally, the blood supply to the temporalis muscle has been compromised or the muscle has been invaded by tumor and it is not available for reconstruction. In these instances, either a myogenous flap such as the pectoralis major or the latissimus dorsi can be used or the area reconstructed with a free flap. The free flap of choice is the radial forearm musculofascial flap. In our series, the complication rate was significantly less in patients whose middle fossa floor was reconstructed with a free flap than when the temporalis muscle is used. The calvarial flap and zygoma are replaced and fixed in place with miniplates. Suction drains are placed and the scalp and facial wound are closed. The tracheostomy is removed once the patient is stable and the lumbar drain pulled if there is no CSF leak at the third or fourth postoperative day.
5.
6. 7.
8.
9. 10.
11.
12.
13.
14.
15. 16. 17.
REFERENCES 1. McLean DH, Buncke HJ Jr. Autotransplant of omentum to a large scalp defect with microvascular revascularization. Plast Reconstr Surg. 1972;49:268. 2. Yassargil MG. Reconstructive Operations on the Cerebral Blood Vessels in Microsurgery Applied to the Neurosurgery. Stuttgart, Germany: Thieme, 1969:96–114. 3. Ketcham AS, Wilkins RN, Van Beuren JM, et al. A combined intracranial facial approach to the paranasal sinuses. Am J Surg. 1963;106:698. 4. Ketcham AS, Chretien PB, Schour L, et al. Surgical treatment of patients with advanced cancer of the paranasal sinuses in neopla-
18.
19. 20.
21.
sia of the head and neck. In: MD Anderson Hospital: Neoplasia of the Head and Neck. Chicago: Year Book Medical Publications, 1974:187. Sisson GA, Bytell DF, Becker SP, et al. Carcinoma of the paranasal sinuses and craniofacial resection. J Laryngol Otol. 1976;90:59. Hybels LR. Venous embolism in head and neck surgery. Laryngoscope. 1980;90:946. Maroon JC, Goodman JM, Hornet TG. et al. Detection of minute venous air emboli with ultrasound. Surg Gynecol Obstet. 1968;127;1236. Broderson BR, Barky N. Acoustic tumor surgery: Anesthetic considerations. In: House WF, Luetje CM, eds. Acoustic Tumors. Vol 2. Baltimore: University Park Press, 1979:3. Fang HS, Ong GB. Direct anterior approach to the upper cervical spine. J Bone Joint Surg (Am). 1962;44-A:1588. Crockard A. Transoral approach to intra/extradural tumors. In: Sekhar LN, Janecka IP, eds. Surgery of Cranial Base Tumors. New York, NY: Raven Press, 1993:225–234. Donald PJ, Bernstein LB. Transpalatal excision of the odontoid process. Trans Am Acad Ophthalmol Otolaryngol. 1978;86:729– 731. Fee WE, Gilmer PA, Goffinet DR. Surgical management of recurrent nasopharyngeal carcinoma after radiation failure at the primary site. Laryngoscope. 1988;98:1220–1226. Fee WE Jr, Moir MS, Choi EC, et al. Nasopharyngectomy for recurrent nasopharyngeal cancer: A 2–17 year follow. Arch Otolaryngol Head Neck Surg. 2002;128(3):280–284. Cocke EW Jr, Robertson JH. Extended unilateral maxillotomy approach. Skull base surgery. In: Donald PJ, ed. Surgery of the Skull Base. 1998:207–237. Wei WI, Lam KH, Sham JS. New approach to the nasopharynx: The maxillary swing approach. Head Neck. 1991;13;200–207. Wei WI. Salvage surgery for recurrent primary nasopharyngeal carcinoma. Crit Rev Oncol Hematol. 2000;124(4):517–521. Janecka IP, Sen C, Sekhar LN, et al. Facial translocation: A new approach to the cranial base. Otolaryngol Head Neck Surg. 1990;103;413–419. Schramm VL. Infratemporal fossa surgery. In: Sekhar LN, Schramm VL, eds. Tumors of the Cranial Base: Diagnosis and Treatment. Vol 24. Mount Kisco, NY: Futura Pub. Co, 1987:421– 437. Fisch U. The infratemporal fossa approaches for nasopharyngeal tumors. Laryngoscope. 1983;93:36–44. Howard DJ, Lloyd G, Lund V. Recurrence and its avoidance in juvenile nasopharyngeal angiofibroma. 2001;111(9):1509– 1511. Binkhorst CD, Schierbeck P, Petten GHW. Neoplasms of the notochord: Report of a case of basilar chordoma and bilateral orbital involvement. Acta Otolaryngol. 1957;47:10.
Chapter 20: Infratemporal/Middle Fossa Tumors 22. Conley J. Concepts in Head and Neck Surgery. New York: Grune & Stratton, 1970:202. 23. Batsakis JH. Soft tissue tumors of the head and neck: Unusual forms. In: Tumors of the Head and Neck-Clinical and Pathological Considerations. Baltimore: Williams & Wilkins, 1974:264. 24. Rice DH, Manusco A, Hanafee WN. Computed tomography with simultaneous contrast sialography. West J Med. 1980;133:321. 25. Stenhouse D, Mason DK. Oral hemangiopericytoma: A case report. Br J Oral Surg. 1968;6:114. 26. Stout AP, Murray MR. Hemangiopericytoma: Vascular tumor featuring Zimmerman’s pericytes. Ann Surg. 1942;116:26. 27. Loke YW. Lymphoepitheliomas of the cervical lymph nodes. Br J Cancer. 1965;19:482. 28. Batsakis JG. Tumors of the Head and Neck—Clinical and Pathological Considerations. Baltimore: Williams & Wilkins, 1974:213. 29. Batsakis JG. Tumors of the Head and Neck—Clinical and Pathological Considerations. Baltimore: Williams & Wilkins, 1974:178. 30. Weiss SW, Goldblum JR. Enzinger and Weiss’s Soft Tissue Tumors. 4th ed. St. Louis, MO: Mosby, 2001. 31. Barnes L. Surgical Pathology of the Head and Neck. Vol 2. 2nd ed. New York: Marcel Decker, 2001. 32. Conley J. Concepts in Head and Neck Surgery. New York: Grune & Stratton, 1970:199. 33. Thompson DE, Forest HM, Hendrick JW, et al. Soft tissue sarcomas involving the extremities and the limb girdles. South Med J. 1971;64:33. 34. Fisher ER, Vuzevski VD. Cytogenesis of schwannoma (neurilemmoma), neurofibroma, dermatofibroma, and dermatofibrosarcoma as revealed by electron microscopy. Am J Clin Pathol. 1968;49:141. 35. Laing D. Nasopharyngeal carcinoma. Otolaryngol Clin North Am. 1969;2:703. 36. Batsakis JG. Carcinomas of the nasopharynx. In: Tumors of the Head and Neck—Clinical and Pathological Considerations. Baltimore: Williams & Wilkins, 1974:123. 37. Toomey JM. Cysts and tumors of the pharynx. In: Paparella MM, Shumrick DA, eds. Otolaryngology. Vol 3. 2nd ed. Philadelphia: WB Saunders, 1980:2323. 38. Coates HL, Pearson GR, Neel HB, et al. Epstein-Barr virus– associated antigens in nasopharyngeal carcinoma. Arch Otolaryngol. 1978;104:427. 39. Yu MC, Ho JH, Lai SH, et al. Cantonese-style salted fish as a cause of nasopharyngeal carcinoma: Report of a case-controlled study in Hong Kong. Cancer Research. 1986;46(2):956–961. 40. Loke YW. Lymphoepitheliomas of the cervical lymph nodes. Br J Cancer. 1965;19:482. 41. Fletcher GH, Million RR. Malignant tumors of the nasopharynx. Am J Roentgenol. 1965;93:44. 42. Vilar P. Nasopharyngeal carcinoma: A report on 24 patients seen over six years. Scot Med J. 1966;11:315. 43. Raventos A, Davis LW. Cancer of the nasal cavity, paranasal sinuses, and nasopharynx: Radiotherapeutic management. In: English GM, ed. Otolaryngology. Vol 3. Hagerstown, MD: Harper & Row, 1979:1–13. 44. Jing GS. Tumors of the nasopharynx. Radiol Clin North Am. 1970;8:323. 45. Rouviere H. Anatomy of the Human Lymphatic System. Translated by MJ Tobias. Ann Arbor: JW Edwards, 1938.
329
46. Donald PJ. Pathophysiology of skull base malignancies. In: Surgery of the Skull Base. Philadelphia: Lippincott-Raven Pub, 1998:62. 47. Donald PJ. Infratemporal fossa-middle cranial fossa approach. In: Surgery of the Skull Base. Philadelphia: Lippincott-Raven Pub, 1998:335. 48. Donald PJ. The significance of invasion of key intracranial structures in skull base surgery for malignancy. Presented at the Triological Society Meeting, Scottsdale, Arizona, May, 1997. 49. Menzek W. Carotid artery assessment and interventional radiologic procedures before skull base surgery. In: Donald PJ, ed. Surgery of the Skull Base. Philadelphia: Lippincott-Raven Pub, 1998:105–118. 50. Ketcham AS, Chretien PB, Schour L, et al. Surgical treatment of patients with advanced cancer of the paranasal sinuses. In: M.D. Anderson Hospital. Neoplasms of the Head and Neck. Chicago: Year Book Medical Publications, 1974:187– 202. 51. Fisch U, Mattox D. Microsurgery of the Skull Base. New York, NY: Thieme Med Pub, 1988:22–24. 52. McCabe FB, Work WP. Parotidectomy with special reference to the facial nerve. In: English GM, ed. Otolaryngology. Vol 4. Hagerstown, MD: Harper & Row, 1967:37. 53. Youmans JR. Neurological Surgery. Vol 2. 2nd ed. Philadelphia: WB Saunders, 1982:854. 54. Proctor B. Anatomy of the Eustachian tubes. Arch Otolaryngol. 1973;97:6. 55. Hardy J, Maina G. Microsurgical anatomy in transsphenoidal hypophysectomy. J Neurol Sci. 1977;21:151. 56. Ritter FN. The Paranasal Sinuses: Anatomy and Surgical Techniques. 2nd ed. St Louis: CV Mosby, 1976:83. 57. Anson BJ, Donaldson JA. Surgical Anatomy of the Temporal Bone and Ear. 2nd ed. Philadelphia: WB Saunders, 1973:9. 58. Anson BJ, Donaldson JA. Surgical Anatomy of the Temporal Bone and Ear. 2nd ed. Philadelphia: WB Saunders, 1973:11. 59. Grant JCB. Grant’s Atlas of Anatomy. 6th ed. Baltimore: Williams & Wilkins, 1972. 60. Pernkoff E. Atlas of Topographical and Applied Human Anatomy. Vol 1. Philadelphia: WB Saunders, 1963. 61. Morgenstein RM. Surgery of the pterygopalatine fossa. In: English GM, ed. Otolaryngology. Hagerstown, MD: Harper & Row, 1976:3. 62. Donald PJ. Sarcomatous degeneration in a nasopharyngeal angiofibroma. Otolaryngol Head Neck Surg. 1979;87:42. 63. Batsakis JG. Tumors of the Head and Neck-Clinical and Pathological Considerations. Baltimore: Williams & Wilkins, 1974: 180. 64. Donald PJ. Skull base surgery for sinus neoplasms. In: Donald PJ, Gluckman JL, Rice DH, eds. The Sinuses. New York: Raven Press, 1995:479. 65. Donald PJ. Skull base surgery for sinus neoplasms. In: Donald PJ, Gluckman JL, Rice DH, eds. The Sinuses. New York: Raven Press, 1995:481. 66. Donald PJ. Skull base surgery for sinus neoplasms. In: Donald PJ, Gluckman JL, Rice DH, eds. The Sinuses. New York: Raven Press, 1995:487. 67. Hybels R. Venous air embolism in head and neck surgery. Laryngoscope. 1980;90:950.
21 Tumors of the Parapharyngeal Space Eric J. Moore and Kerry D. Olsen
and contains attachments of the stylopharyngeal aponeurosis. The inferomedial wall continues with fascia that joins the styloglossus and stylopharyngeus muscles. The parapharyngeal space has been further divided by most authors into a prestyloid and a retrostyloid compartment (Fig. 2). Fascia that extends from the styloid process to the tensor veli palatini muscle crosses posteriorly in the parapharyngeal fat and separates the parapharyngeal space into these two areas. More posteriorly, this fascial plane blends with the styloid muscles. The prestyloid space extends superiorly into a blind pouch formed by the joining of the medial pterygoid fascia to the tensor veli palatini fascia. This space contains a variable portion of the retromandibular deep lobe of the parotid gland. In addition, a small branch of the fifth cranial nerve crosses this area to reach the tensor veli palatini muscle. Most of the prestyloid parapharyngeal space is composed of fat. Therefore, tumors in this area are generally limited to salivary lesions, lipomas, and rare neurogenic tumors. ( Table 1) The retrostyloid compartment or poststyloid compartment contains the carotid artery and jugular vein located posterolateral to the artery at the skull base. Cranial nerves IX through XII accompany these vessels, with the 10th nerve occupying a position between the artery and the vein. The 11th nerve crosses the vein anteriorly or posteriorly, and the 9th nerve crosses the carotid artery laterally. The 12th nerve ends its vertical course outside the parapharyngeal space. This compartment also contains the sympathetic chain, lymph nodes, and glomus tissue. These structures all serve as a potential source for a retrostyloid tumor. The retropharyngeal space is separated from the retrostyloid space by a thin fascial layer that is a minimal barrier to the spread of tumor or infection. The fascia uniting the styloid process to the mandibular ramus is called the stylomandibular ligament. This ligament forms one of the boundaries of the stylomandibular tunnel, as described by Patey and Thackray (2). The remaining borders include the skull base and ascending ramus of the mandible. Extension of tumors through the rigid opening of the stylomandibular tunnel will often be noted by constricted tumor growth in this narrow area. Surgical entry to the parapharyngeal space is improved by dividing the stylomandibular ligament, removing the styloid process, or dislocating the jaw forward. The parapharyngeal space has numerous lymphatics that drain the paranasal sinuses, for example, the oropharynx, the oral cavity, and a portion of the thyroid gland. These nodes are connected superiorly to the node of Rouviere, situated in the retropharyngeal space, which drains the nasopharynx, upper oropharynx, and sinuses. Inferiorly, the lymphatic drainage continues to the jugular digastric nodes, which are outside and inferior to the parapharyngeal space. A direct
SURGICAL ANATOMY The anatomy of the parapharyngeal space is either wonderfully intricate or frighteningly complicated, depending on the experience of the operator. Failure to appreciate the anatomic relationships can lead to selection of an incorrect surgical approach. The result may be inadequate access, with difficulty with tumor removal, damage to vital structures, or tumor spillage, and thus recurrent neoplasms. The parapharyngeal space is often described as an inverted pyramid with its base at the skull and apex at the greater cornu of the hyoid bone (Fig. 1). The parapharyngeal space is further compartmentalized by thick fascial layers that direct tumor growth. Prior descriptions of these fascial layers have varied (1). The superior border of the parapharyngeal space is a small portion of the temporal bone (Fig. 2). The superomedial wall is enclosed by a fascial connection from the medial pterygoid plate to the spine of the sphenoid. This fascia passes medial to the foramen ovale and the foramen spinosum. These foramina are not included in the superior limits of the parapharyngeal space but, rather, are in the infratemporal fossa or masticator space. The inferior boundary of the parapharyngeal space ends at the junction of the posterior belly of the digastric muscle and the greater cornu of the hyoid bone. The firm fascial attachments in this area limit parapharyngeal space extension inferior to the hyoid bone. This fascia, however, can be weak and may be an ineffective barrier to the spread of infections. The posterior border of the parapharyngeal space is formed by the fascia over the vertebral column and paravertebral muscles. The anterior limit is composed of the pterygomandibular raphe and medial pterygoid fascia. The lateral wall of the parapharyngeal space is made up of the fascia overlying the medial pterygoid muscle and the ramus of the mandible. The fascia of the medial pterygoid muscle superiorly incorporates the sphenomandibular ligament that extends from the spine of the sphenoid to the lingula of the mandible. This dense fascia then continues as a firm layer to the skull base. This fascial layer separates the parapharyngeal space from the inferior alveolar nerve, the lateral pterygoid muscle, and the condyle of the mandible. The retromandibular portion of the deep lobe of the parotid gland also forms a small portion of the lateral border, as does a portion of the posterior belly of the digastric muscle. Superiorly, the medial border is formed by the approximation of the fascia from the tensor veli palatini muscle to the medial pterygoid muscle. The pharyngobasilar fascia forms the posteromedial border of the retrostyloid space near the levator palatini muscle. Inferiorly, the medial border is contiguous with the fascia over the superior constrictor muscle, 331
332
Moore and Olsen
connection from the lymphatics of the thyroid gland to the parapharyngeal nodes was reported (3). Lymph channels extend along the posterior wall of the pharynx and terminate in the lateral retropharyngeal node. This pathway exists as a result of the embryonic development of the thyroid gland from the tongue base. Similarly, hypopharyngeal and laryngeal tumors that involve the posterior pharyngeal wall may also metastasize to the lateral retropharyngeal nodes. In summary, the only nonrigid borders of the parapharyngeal space are the medial and inferior areas. In addition, laterally, a small opening occurs in the stylomandibular tunnel. It is in this last area that the more medial fascia of the tensor veli palatini muscle separates from the more lateral fascia of the medial pterygoid muscle, causing a space for parotid tumors to enter the prestyloid area.
REGIONAL PATHOLOGY AND DIFFERENTIAL DIAGNOSIS
Figure 1 Inverted pyramidal shape of the parapharyngeal space.
Tumors of the parapharyngeal space include primary neoplasms, direct extension from adjacent regions, and metastatic disease (Table 2). Malignant tumors can invade the parapharyngeal space from the nasopharynx, oropharynx, mandible, maxilla, oral cavity, or parotid gland. The node of Rouviere is often the first site of metastasis of nasopharyngeal or antral carcinoma (4). Primary parapharyngeal tumors can extend intracranially through the jugular foramen or into the retropharyngeal space. Considering the complexity of the contents of the parapharyngeal space, it is not surprising that there are a wide variety of neoplastic tumors. Reviews of parapharyngeal space tumor series demonstrate that the majority of neoplasms in this area are of salivary gland origin, followed by neurogenic, lymphatic, and metastatic lesions (5).
Figure 2 The superior border of the parapharyngeal space is formed by the base of skull. Note the tensor veli palatine fascia that separates the parapharyngeal space into a prestyloid and poststyloid compartment.
Chapter 21: Tumors of the Parapharyngeal Space Table 1 Contents of Prestyloid and Poststyloid Compartments of the Parapharyngeal Space Structure
Prestyloid compartment
Poststyloid compartment
Vessels
Internal maxillary artery Ascending pharyngeal artery Auriculotermporal nerve
Internal carotid artery Internal jugular vein
Nerve Other
Lymph nodes Deep lobe parotid gland Ectopic salivary rests
Cranial nerves IX,X,XI,XII Cervical sympathetic chain Lymph nodes Glomus bodies
Salivary Gland Neoplasms In the majority of reports, pleomorphic adenoma is the most common parapharyngeal space tumor. This tumor generally originates from the deep lobe of the parotid gland, but also can occur from extraparotid salivary tissue. The most common deep-lobe parotid tumors arise from the gland deep to the facial nerve, yet remain lateral to the mandible. These tumors, in the majority of cases, do not involve the parapharyngeal space. However, a deep-lobe parotid tumor may have an external component palpable anterior to the tragus and have a pharyngeal component that extends through the stylomandibular tunnel into the parapharyngeal space. This “dumbbell” tumor may have a variable portion of the mass in the parapharyngeal space. Pleomorphic adenomas can also arise from the retromandibular portion of the parotid gland. Medial extension occurs where the external carotid pierces the parotid fascia inferior to the stylomandibular ligament. The tumor expands into the parapharyngeal space and displaces the tonsil and palate. The typical tumor configuration is more round. As these tumors grow and enlarge, they may also become palpable below the angle of the jaw. They do not expand laterally to present as a pretragal mass, as do dumbbell tumors (6). These tumors can enlarge to such a degree that they displace the soft palate and obstruct the nasopharynx. They cause surprisingly Table 2
few symptoms and generally have no palpable component in the neck. A final origin of parotid parapharyngeal space tumors is the tail of the superficial lobe. These tumors can form round lesions that grow medially and cranially to present as a parapharyngeal space mass. The pharyngeal component is generally the largest, but the tumors also have an external palpable component situated posterior and inferior to the angle of the jaw. Extraparotid salivary tissue is also a source of parapharyngeal space neoplasms. These lesions are usually benign, and they are usually pleomorphic adenomas. These tumors arise in ectopic salivary rests in lymph nodes or from ectopic salivary gland lateral to the superior pharyngeal constrictor muscle. This salivary tissue lying in the parapharyngeal fat may be the source for parapharyngeal space tumors with no obvious connection to the deep lobe of the parotid gland and no extension through the constrictor muscle. Parapharyngeal salivary tumors also may arise from serous glands located in the pharyngeal mucosa medial to the superior constrictor muscle (7). Other benign tumors of salivary origin have been reported in the parapharyngeal space. These include Warthin tumors, oncocytomas, and benign lymphoepithelial lesion. The frequency of malignant parapharyngeal salivary tumors compared to benign salivary tumors is approximately 1:3 (5). Malignant tumors of salivary origin include mucoepidermoid carcinoma, adenoid cystic carcinoma, acinic cell carcinoma, malignant mixed carcinoma, squamous cell carcinoma, adenocarcinoma, and a case of a malignant Warthin tumor (8).
Neurogenic Tumors The most common neurogenic neoplasm found in the parapharyngeal space is the neurilemmoma (schwannoma). The site of origin is generally the vagus nerve or sympathetic chain. Overall, approximately 30% of all neurilemmomas occur in the head and neck area, with the majority originating
Neoplasms of the Parapharyngeal Space Benign
Malignant
Salivary gland
Pleomorphic adenoma Warthin tumor Oncocytoma Benign lymphoepithelial lesion
Neurogenic
Neurilemmoma Neurofibroma Vagal paraganglioma Carotid body tumor Lymphatic malformation Hemangioma Vascular malformation Dermoid tumor Meningioma Rhabdomyoma Teratoma Lipoma Branchial cleft cyst Hibernoma Leiomyoma Hemangioepithelioma
Adenoid cystic carcinoma Acinic cell carcinoma Adenocarcinoma Mucopeidermoid carcinoma Carcinoma expleomorphic adenoma Neurofibrosarcoma
Other
Metastatic
333
Lymphoma Rhabdomyosarcoma Plasmacytoma Chordoma Fibrosarcoma Fibrous histiocytoma Hemangiopericytoma Liposarcoma
Squamous cell carcinoma Thyroid carcinoma Meningioma
334
Moore and Olsen
from the vagus nerve. The vagus nerve has been reported to be the nerve of origin in 50% of parapharyngeal neurilemmomas (9). The most common tumor of the vagus nerve, however, is a paraganglioma, followed by neurilemmoma, neurofibroma, and neurofibrosarcoma. Cranial nerves IX through XII and the sympathetic chain are all encased in Schwann cells and can give rise to neurilemmomas. The cervical sympathetic chain is the second most common nerve of origin of the parapharyngeal space. Preoperative Horner syndrome is uncommon. The glossopharyngeal nerve and hypoglossal nerve are rarely the source of parapharyngeal neurogenic tumors. When large parapharyngeal neurilemmomas involve adjacent nerves, it can sometimes be difficult to determine the true nerve of origin. Neurilemmomas of the parapharyngeal space are generally benign and slow-growing neoplasms. They rarely cause neuropathy of their nerve of origin. They frequently present as neck masses or incidentally noted lesions on imaging. With time, they can cause obstructive symptoms or loss of function. Removal of neurilemmomas can be accomplished with preservation of the nerve of origin and potential return of function. Occasionally, the nerve will have to be sacrificed, resulting in permanent dysfunction. Paragangliomas that involve the parapharyngeal space originate from the vagal or carotid bodies. Glomic tissue of neural-crest origin has been found around the surface of the nodose ganglion, and vagal paragangliomas are most commonly found in the parapharyngeal space. Carotid body tumors that extend into the parapharyngeal space are more rare. They can present as a painless mass below the angle of the jaw that extends above the posterior belly of the digastric muscle into the parapharyngeal space. Paragangliomas can be multicentric 10% of the time, and bilateral imaging of the parapharyngeal space is indicated. Patients with multiple paragangliomas or a family history of paraganglioma have, by definition, the familial form of paraganglioma. The gene for paraganglioma has been sequenced, and family members of patients with affected individuals can be screened for the genetic mutation (10). Many times the multiple tumors are occult, and screening and imaging is necessary for detection. Neurofibromas are the third most common neurogenic tumor in the parapharyngeal space. Neurofibromas generally occur as multiple lesions—they intimately involve the nerve, and removal of tumor with preservation of the nerve is not possible. Cranial nerve deficits occur with neurofibromas more often than with schwannomas. Von Recklinhausen disease represents an autosomal dominant disorder characterized by five or more caf´e au lait cutaneous lesions and multiple neurofibromas. Malignancy occurs in about 10% of patients with von Recklinhausen disease. Although cutaneous neurofibromas are the most common manifestations, cranial nerves can be involved with neurofibromas, and they may present in the parapharyngeal space. Several malignant nerve tumors, such as malignant paragangliomas and schwannosarcomas, occur in the parapharyngeal space. Malignant neuroblastoma has been reported, as has malignant solitary schwannoma (11).
Miscellaneous Tumors Other rare tumors have been found in the parapharyngeal space. Temporal bone meningiomas may arise from extracranial extension of a primary intracranial tumor, a neoplasm arising in the jugular foramen, or a tumor originating from arachnoid cell clusters within the trunk of a cranial nerve
of from its peripheral sheath. Metastasis can occur from a primary intracranial meningioma. Numerous vascular lesions were reported as primary neoplasms of the parapharyngeal space. These include hemangiomas and lymphatic malformations, arteriovenous malformations, venous malformations, hemangiopericytoma, and internal carotid aneurysms. Other lesions include lipomas, hibernoma, rhabdomyoma, teratoma, branchial cleft cysts, and dermoid tumors. Miscellaneous malignant tumors are also found in the parapharyngeal space. The most commonly reported lesion is isolated lymphoma. Chordoma can spread to or even originate in the parapharyngeal space, and malignant fibrous histiocytoma had been reported in this region. Additional unusual malignant tumors originating in the parapharyngeal space include malignant teratoma, chondrosarcoma, rhabdomyosarcoma, malignant hemangiopericytoma, meningiosarcoma, fibrosarcoma, and plasmacytoma (5). Squamous cell carcinoma can metastasize to lymph nodes in the parapharyngeal space. This is most common with nasopharyngeal tumors, oral cavity carcinoma, oropharyngeal carcinoma, and hypopharyngeal lesions. It is rare to have a parapharyngeal metastasis present as the original manifestation of a tumor. This can occur with papillary thyroid carcinoma, follicular thyroid cancer, and medullary thyroid cancer.
CLINICAL ASSESSMENT Tumors of the parapharyngeal space most often present as an asymptomatic mass in the neck or oropharynx. They are often discovered on routine physical examaniation. Their presence should be suspected when a subtle fullness is noted in the soft palate or tonsillar region, or when there is mild fullness near the angle of the jaw. Clinical detection of early parapharyngeal lesions may be difficult. The tumors must often grow to 3.0 cm in size before they can be palpated. Small tumors in the parapharynx cause few symptoms. As the tumors enlarge and extend superiorly, they may cause symptoms related to the e eustachian tube. As tumors expand medially, voice change, nasal obstruction, aspiration, and dyspnea may occur. Rarely, tumors have been found that require immediate tracheotomy for relief of upper airway obstruction. As tumors enlarge, they may compress the 9th, 10th, 11th, or 12th cranial nerve, causing hoarseness, dysphagia, and dysarthria. Horner syndrome may also be produced by tumor pressure on the superior cervical ganglion. For a benign tumor to cause significant nerve deficits, it must enlarge to a considerable degree. Pain, trismus, or cranial nerve palsy often suggests malignancy. Obstructive sleep apnea symptoms may result from parapharyngeal space lesions. The symptoms include progressing onset of snoring, restless sleep, and daytime somnolence. Airway encroachment may become significant as the tonsillar fossa medial displacement becomes pronounced. Malignant tumors of the parapharyngeal space can cause carotid sinus hypersensitivity and glossopharyngeal neuralgia. Asystole, bradycardia, and hypertension have been reported (12). These symptoms may be due to neural irritation of the glossopharyngeal afferent fibers. Tumors of the parapharyngeal space are often misdiagnosed as infections or tonsil tumors. Patients often complain of a mild sore throat or globus sensation, and sometimes they complain of dysphagia. The swelling in the tonsillar and soft palatal region may be misdiagnosed as a peritonsillar abscess.
Chapter 21: Tumors of the Parapharyngeal Space
Delays in diagnosis have also occurred because patients were being treated for presumed nasal obstruction, eustachian tube dysfunction, or serous otitis media. Patients have been misdiagnosed with temporomandibular joint pathology when they actually had a parapharyngeal space lesion. The performance of a complete head and neck examination is one of the most important aspects of the evaluation of parapharyngeal space tumors. However, the anatomic location of the parapharyngeal space makes it difficult to accurately assess tumor presence and size. For a mass to be noted on clinical examination, it usually must attain a diameter of 3 cm. Only a mass of considerable size will cause a visible bulge or palpable abnormality of the lateral pharyngeal wall or external neck. Swelling of the medial wall of the oropharynx is generally the first sign of a parapharyngeal space lesion. Parotid tumors displace the tonsil, and neurogenic tumors often displace the posterior portion of the pharynx and posterior tonsillar pillar. As tumors enlarge superiorly, they fill the space between the heads of the tensor veli palatini muscle, causing soft palate and nasopharyngeal swelling. As a tumor extends inferiorly, it presents as a palpable mass near the angle of the jaw. Cranial nerve function should be noted. Bimanual palpation with one of the physician’s fingers in the patient’s mouth and the physician’s other hand on the patient’s neck assesses mobility, pulsation, and tumor extent. This may help determine the origin and extent of the tumor. The finding of a pulsatile mass is generally not a helpful differentiating sign, as many of the tumors will transmit carotid pulsations. If a preauricular mass is noted at physical examination in addition to a lateral pharyngeal mass, this indicates a parotid tumor extending through the stylomandibular tunnel. The presence of cervical lymphadenopathy may indicate malignancy. Flexible nasal endoscopy can help determine the inferior extent of tumor, and it can also lend information regarding function of the 9th and 10th cranial nerves. The status of the patient’s airway should be assessed, and a determination of the expected difficulty of orotracheal intubation should be noted.
DIAGNOSTIC IMAGING Radiographic study of all parapharyngeal space tumors is essential. A computed tomographic (CT) scan with and without contrast medium or a magnetic resonance imaging (MRI) study with gadolinium should be performed in all cases. Angiographic procedures may also be necessary in select cases. The results of these studies aid significantly in diagnosis and treatment planning. Computed tomography imaging is capable of displaying the soft tissues of the parapharyngeal space extremely well. One of the most important features in radiographic evaluation of parapharyngeal space lesions is to assess whether they lie anterior or posterior to a plane from the styloid process to the medial pterygoid plate. This is the plane of the fascia of the tensor veli palatini muscle which divides the parapharyngeal space into a prestyloid and poststyloid compartment. A lesion in the prestyloid compartment will be anterior to the carotid artery and posterior to the medial pterygoid muscle (13) (Fig. 3). In general, prestyloid tumors are usually salivary gland neoplasms that displace the carotid sheath contents posteriorly. Poststyloid tumors are normally of neurogenic or vascular origin. Retrostyloid tumors displace the internal carotid artery in an anteromedial direction.
335
The exception to this rule is schwannomas, which can displace the carotid artery in different directions because of the unpredictable tumor position between the great vessels and the site of origin of the tumor. Vessel displacement depends on the nerve of origin and whether the tumor arises near the base of the skull or in the inferior portion of the parapharyngeal space. CT scans may be helpful in separating prestyloid deeplobe parotid tumors from extraparotid salivary neoplasms. The best way to distinguish between these two lesions is the finding of a fat plane between the deep lobe of the parotid gland and the posterolateral aspect of a mass. The fat represents compressed fibrofatty supporting tissues in the parapharyngeal space. When seen, the tumor is extraparotid. Unfortunately, for lesions larger than 4 cm, the fat plane is obliterated and it is difficult to determine whether a tumor originates in a minor salivary gland or the parotid gland. It is important to look for evidence of skull base or cervical vertebral erosion and extension through the jugular foramen into the cranial cavity also. Benign prestyloid tumors can cause erosion of the pterygoid plate, and this finding is not pathognomonic for malignant lesions. Radiographically, lowgrade malignancies are difficult to distinguish from benign parapharyngeal space tumors. Large pleomorphic adenomas have a less homogeneous appearance on CT and contain irregular areas of minimal enhancement, which can give an appearance similar to that of many neuromas. Neurilemmomas also often have areas of hemorrhage, cystic necrosis, and fatty deposition (14). Lesions that show enhancement on CT with contrast medium include paragangliomas, hemangiomas, hemangiopericytomas, aneurysms, and neurilemmomas. Indications of malignancy in the parapharyngeal space include irregular tumor margins, spread into surrounding tissues and fat planes on CT, and evidence of enlarged necrotic nodes in the cervical area or in the retropharyngeal space. CT scan has been shown to be helpful in assessing the parapharyngeal space as an area of extension of nasopharyngeal carcinomas. The parapharynx is often involved directly by nasopharyngeal tumors, and CT aids in the assessment of this region in patients who have nasopharyngeal carcinoma. MRI also provides useful preoperative information about the extent of tumor and its relationship to surrounding structures. The capability to image the coronal and sagittal
Figure 3 Axial CT scan of patient with left parapharyngeal pleomorphic adenoma. Note that lesion is anterior to the internal carotid artery and posterior to the medial pterygoid muscle.
336
Moore and Olsen
Figure 4 T1-weighted MRI of same patient as Figure 3 demonstrating left parapharyngeal pleomorphic adenoma.
Figure 6 Sagittal MRI of same patient, note extension of within 1 cm of base of skull and encasement of internal carotid artery.
planes directly with MR scan is a significant advantage over CT. MRI also has the ability to distinguish between different types of lesions based on their difference in signaling an enhancement. Pleomorphic adenomas, for instance, has low signaling on T1-weighted images and bright signaling on T2-weighted images, and paragangliomas often demonstrated flow voids within areas of high vascularity (Figs. 4–6). The presence of flow voids on imaging of a poststyloid lesion can suggest paraganglioma over less vascular schwannomas. This information can help in counseling and preoperative planning. It is difficult to distinguish between malignant and benign parapharyngeal lesions based on images, but some findings may suggest malignancy. Irregular borders and obliteration of fat planes, particularly around nerves, can suggest malignancy. Invasion of the skull base and dura is often an indicator of malignancy, as is the presence of lymphadenopathy. If on the basis of history, physical examination, and imaging studies a paraganglioma is suspected, then patients should undergo angiography. Angiography can help confirm the diagnosis, as in the case of carotid body tumors, if a classic lyre sign with separation of the internal and external carotid artery is present. It can also allow identification of feeding vessels, and simultaneous embolization can be performed
(Figs. 7 and 8). Embolization is used by the authors for extensive paragangliomas extending to the skull base and for some carotid body tumors. If embolization is planned, it should be performed within 24 hours of the definitive operative therapy to prevent inflammatory reaction from complicating tumor removal and to prevent revascularization. Prior to surgery, it may also be necessary to perform a carotid occlusion study during the angiogram to determine if the patient can tolerate loss of the carotid artery. This must always be done whenever a malignant tumor involves the retrostyloid portion of the parapharyngeal space or for extensive vascular tumors that surround the carotid artery at the base of skull. A cerebral perfusion agent is injected intravenously with a balloon inflated in the carotid artery. The balloon is then deflated and a single photon-emission CT of the brain is obtained. If there is evidence of hypoperfusion of the ipsilateral cerebral hemisphere, a repeat scan can be subsequently performed. This helps determine if hypoperfusion was due to temporary balloon occlusion or to a preexisting
Figure 5 Axial Postgadolinium MRI of patient with poststyloid vagal paraganglioma. Note flow voids within the tumor.
Figure 7 Angiogram of vagal paraganglioma shown in Figures 5 and 6. The tumor displaces the carotid vessels anteriorly and fills from a tributary of the ascending pharyngeal artery.
Chapter 21: Tumors of the Parapharyngeal Space
Figure 8 Angiogram postembolization of carotid vessel.
abnormality. During the study, the patient also undergoes electroencephalographic monitoring and is observed for clinical evidence of neurologic dysfunction. If patients fail balloon occlusion study, then preparations for operative carotid reconstruction may be necessary. Ultrasound is rarely indicated for assessment of tumors of the parapharyngeal space, with the exception of possible utilization to aid in fine-needle aspiration biopsy. Positron emission tomography may be useful in assessment of malignant tumors in some circumstances.
PREOPERATIVE PREPARATION Biopsy of tumors of the parapharyngeal space is rarely necessary after careful history, physical examination, and imaging. Transoral biopsy is contraindicated since the technique causes the pharyngeal mucosa to adhere to the tumor capsule, making subsequent removal more difficult. In addition, there is a risk of tumor spillage, hemorrhage, and other untoward complications that may be difficult to manage. If biopsy is necessary, fine-needle aspiration can easily be performed either directed by palpation or with the assistance of ultrasound or CT guidance. Fine-needle biopsy may be useful in confirmation of malignancy to help with treatment planning and patient counseling. Fine-needle aspiration is generally accurate in diagnosis of pleomorphic adenoma and other tumors of salivary origin. It may be less useful in neurogenic tumors, which often yield a hypocellular aspirate. Not every tumor of the parapharyngeal space deserves surgical extirpation. As in every surgical consultation, the surgeon needs to weigh the benefits of removal against the morbidity of the procedure to adequately advise the patient. Removal of paraganglioma of neural origin is nearly always associated with permanent disability of the involved nerve. Some authors have questioned the benefit of removal of paragangliomas of the base of skull (15). Similarly, removal of neurilemmomas arising from major cranial nerves in the parapharyngeal space can result in morbidity such as dysphagia and dysphonia. These neural tumors often grow slowly, and some reports demonstrate good growth control with nonoperative therapy such as external beam irradiation
337
(16). In general, we recommend operative therapy of prestyloid parapharyngeal space lesions. These tumors often arise from salivary gland origin, and the morbidity of removal of these lesions is usually low. The benefit of definitive diagnosis and avoidance of long-term morbidity associated with growth and malignant potential outweighs potential complications in most of these patients. The counseling regarding removal of paragangliomas and parapharyngeal space schwannomas is heavily influenced by the age of the patient. Patients younger than 70 years, who are acceptable surgical candidates based on other comorbidities, are usually offered surgery. Patients older than 70 years are commonly counseled regarding serial imaging and selection for removal based on evidence of growth, malignant potential, and potential morbidity of observation. If on the basis of workup a paraganglioma is suspected, the patient should undergo testing for possible catecholamine secretion. Glomus jugulare tumors, carotid body tumors, and vagal paragangliomas can all secrete catecholamines. Historical points that may indicate the presence of catecholamine production include labile hypertension, tremulousness, headache, palpitations, and sweating. Urine should be collected for the presence of vanillylmandelic acid, metanephrines, dopamine, epinephrine, and norepinephrine. Serum catecholamines should also be quantitated. Failure to discover a secreting tumor preoperatively can have dire consequences during surgical removal of the tumor. If a secreting tumor is found, preoperative blockade with propranolol and phenoxybenzamine may help control intraoperative arrhythmias and hypertension. Following the complete evaluation, the surgeon should thoroughly discuss with the patient the assessment, treatment plan, and goals of surgery. This plan should be individualized to the specific patient and the tumor characteristics. The discussion includes an assessment of benign versus malignant tumor behavior, but even in the setting of fine needle aspiration (FNA) diagnosis, the surgeon should prepare the patient for an alteration of the diagnosis based on intraoperative findings and frozen-section analysis. The treatment plan must be designed around safe and complete tumor extirpation, avoidance of piecemeal removal, appropriate management of cervical lymphatics, nerve preservation if oncologically sound, and appropriate rehabilitation measures as necessary. All patients undergoing parapharyngeal surgery should be counseled regarding incisional placement and scar; expected cosmetic defect; expected postoperative function and time to full recovery; possibility of need for additional therapy as indicated by tumor type; and postoperative sequelae, such as paresthesias, dysphagia, dysphonia, and first bite pain. The necessity of thorough preoperative counseling in this operation cannot be overestimated, and a carefully planned and executed operation can still result in a dissatisfied surgeon and patient if the patient has not been informed as to appropriate expectations after the operation.
SURGICAL TECHNIQUE Although many surgical approaches have been described for the treatment of parapharyngeal lesions, the vast majority of tumors in this area can be treated by the cervical– parotid approach with or without minor variations. The four most commonly applied operations are (1) the cervical, (2) the cervical-parotid, (3) the cervical-parotid with a lateral or midline mandibulotomy, and (4) the transoral. The transoral approach has been reported to be effective for removing
338
Moore and Olsen
select benign, minor salivary tumors that originate high in the parapharyngeal space (17). However, this approach gives poor exposure, lack of access to regional vessels and nerves, and high risk of tumor spillage with possible recurrence. An incorrect approach can result in significant morbidity and mortality. Our most commonly used surgical approaches for parapharyngeal lesions are the cervical–parotid approach and the cervical–parotid approach with midline mandibulotomy. The cervical–parotid approach is used for the deep-lobe parotid neoplasms, extraparotid salivary tumors, and most neurogenic tumors. This approach can be combined with a craniotomy for tumors that extend intracranially. Intracranial extension is then treated at the same operation. The cervical approach with midline mandibulotomy is used for highly vascular tumors that extent into the superior parapharyngeal space and for tumors confined to the superior parapharyngeal region, that is, the eustachian tube and skull base. This procedure is indicated for tumors that have invaded the skull base or vertebral bodies. In these cases and in those with obvious intracranial extension, the operation is performed in conjunction with a neurosurgeon.
Positioning Before entering the operating theater, the surgeon should arrange for the availability of a pathologist to provide reliable frozen-section review of the tissue. This will save many patients the cost, time, and morbidity of returning for further surgical resection of an undiagnosed malignancy. A quality assurance system to insure correct patient and site of surgery should be in place. Appropriate imaging studies should be thoroughly reviewed and visible during the operation. The risk of bleeding is low, and patients do not generally need to be typed and cross-matched for blood transfusion during the operation. An exception to this would be in the case of paragangliomas, when blood availability should be ensured. Antibiotics are not routinely administered unless there are preoperative signs of infection. Intraoperative EEG monitoring or cranial nerve monitoring is performed only in the case of carotid body tumors and skull base paragangliomas. The operation is performed under general endotracheal anesthesia with the endotracheal tube positioned and taped to the oral commissure and cheek opposite to the lesion. The patient is placed in a 45-degree reverse-Trendelenburg position or “lounge chair position” so that the head is higher than the heart. The head is turned to the opposite side of the lesion, and the neck is extended by placement of a rolled sheet under the shoulders. The patient is then prepared by sterile scrub and draped so that the ear, lateral corner of the ipsilateral eye, ipsilateral oral commissure, and entire ipsilateral neck are visible in the field. The surgeon stands on the side of the patient ipsilateral to the gland to be dissected, the assistant stands at the head and opposite the surgeon, and the scrub tech stands on the side of the surgeon.
Incision The placement of the incision is marked with a surgical marker. The incision begins in the preauricular crease at the superior root of the helix. The incision then descends in a preauricular crease, curves gently below the lobule, and then turns anteriorly to run horizontally in a skin crease approximately 2 fingerwidths below the angle of the mandible. This limb of the incision should be oriented so that it could be extended into an incision that will accommodate dissection of the neck, or so that it could be extended up through the submental and lip area to incorporate mandibulotomy. The
incision should not extend far posteriorly into the thin skin below the lobule over the mastoid tip. This skin will become ischemic in patients with tobacco abuse or diabetes, and the surgeon will risk developing skin flap loss. The skin incision can be combined with a postauricular incision to perform mastoidectomy or craniotomy for tumors that extend intracranially, via the jugular foramen. Intracranial extension is removed at the same time as the parapharyngeal tumor, not as a later procedure. The incision can be placed in the retrotragal area to hide the scar better in patients who are particularly concerned with scar camouflage. If the surgeon performs this type of incision, then the incision line should lie just inside the anterior edge of the tragus, and the dissection of the skin over the tragus should be very thin, in the immediate subcutaneous plane to avoid elevation of the tragal perichondrium. The surgeon may crosshatch the incision lines superficially with a no. 10 or 15 blade to assist in precise realignment during closure. The incision is then performed from superior to inferior through the skin into the subcutaneous tissue with the scalpel. Double-skin hooks are placed into the facial flap, and the flap is raised for 1 cm with the blade. The assistant places firm upward retraction on the skin flap with the hooks, and back-traction is applied on the skin with a sponge either by the surgeon or by a second assistant.
Soft Tissue Dissection The surgeon should raise the flap immediately over the parotid fascia, which is recognizable as a white fibrous layer deep to the subcutaneous fat and superficial musculoaponeurotic system layer. The neck flap is dissected in the subplatysmal plane to expose the sternocleidomastoid muscle, submandibular gland, and anterior and inferior neck. The inferior portion of the parotid gland is separated sharply from the sternocleidomastoid muscle, the posterior belly of the digastric muscle, and cartilaginous ear canal. The main trunk of the facial nerve is identified. If one is dealing with a deeplobe parotid tumor, a superficial prostatectomy is performed. If the tumor is not in the deep lobe of the parotid gland, then the facial nerve trunk and inferior division is isolated and retracted gently superiorly (Fig. 9). The sternocleidomastoid muscle is retracted laterally, and the accessory nerve can be identified. The posterior belly of the digastric muscle is isolated completely down to its insertion on the hyoid bone. The dense stylomandibular fascia between the inferior parotid gland and the submandibular gland is divided so that the submandibular gland can easily be retracted medially for exposure. The posterior bellies of the digastric muscle and the stylohyoid muscle are separated from their attachments at the mastoid tip and from the styloid process, and are retracted medially. This gives further superior exposure of the internal carotid artery, jugular vein, and adjacent nerves. The external and internal carotid arteries, common carotid artery, and internal jugular vein are isolated, and vessel loops are placed around the internal and external carotid arteries for security. Upper jugular nodes may be removed to aid in exposure, and cranial nerves X and XII are isolated. The external carotid artery is now easily seen passing into parotid tissue in front of the styloglossus muscle. This artery and its corresponding vein are divided. The angle of the mandible is retracted anteriorly if necessary for exposure— paralysis of the patient may help at this point. This stretches the stylomandibular ligament so that it can easily be palpated and divided. Dividing the ligament creates a wide opening into the parapharyngeal space (Fig. 10). The tumor may be able to be visualized and palpated at this point.
Chapter 21: Tumors of the Parapharyngeal Space
Figure 9 Exposure and mobilization of the inferior division and trunk of the facial nerve.
Bony Dissection and Osteotomies Although rarely necessary for parapharyngeal tumors, the cervical approach with midline mandibulotomy can be used for the removal of highly vascular tumors that extend to the superior portion of the parapharyngeal space. Tumors that are confined solely to the superior parapharynx are also approached in this way. The procedure is also indicated for tumors that have invaded the skull base or vertebral bodies. A mandibulotomy should be in the informed consent discussion for any patient with a parapharyngeal tumor that extends to the skull base or where bleeding and/or poor exposure are anticipated. A tracheostomy is performed first, because subsequent edema can compromise the airway. The incision is the same as for the cervical parotid approach, and it is extended up through the submental incision, through or around the mentum, and into the midline lower lip (Fig. 11). The digastric muscle is freed from the hyoid bone, and the digastric tendon and anterior belly are retracted superiorly with the submandibular gland. A plating system with an inferior border plate and a smaller upper plate is bent, applied, drilled, and then set aside for subsequent closure. The mandible is then divided with a sagittal saw in a stair-step manner in the midline, preserving the incisor teeth if good dentition is present. An incision is made intraorally along the floor of the mouth
339
Figure 10 Division of the stylomandibular fascia to open the PPS.
extending this behind the anterior tonsillar pillar and onto the hard palate. The submandibular duct is included with the soft tissue of the mandible and the lingual nerve is preserved as it stretches across the operative field. The hypoglossal nerve is followed to its entrance into the tongue and forms the lower border for division of the musculature of the tongue (Fig. 12). The supporting musculature of the floor of the mouth is divided as are the styloglossus and stylopharyngeus muscles where they enter the pharynx and tongue. A plane is then established lateral to the constrictor muscles and the parapharyngeal space is widely exposed (Fig. 13). If the tumor involves the region of the nasopharynx, the muscles of the eustachian tube can be divided. The retropharyngeal space is entered and the contents of the nasopharynx can be retracted medially. Tumor removal can safely be done by following the carotid intracranially if necessary.
Tumor Resection Deep-lobe parotid tumors can usually be palpated and removed with blunt and digital dissection (Fig. 14). It is not necessary to perform a superficial or total parotidectomy for most deep-lobe parotid tumors. In our experience, most deep-lobe parotid tumors are not the dumbbell type, but are rounded and originate from the retromandibular portion of the deep lobe of the parotid gland. There often is a narrow attachment of parotid tissue to the tumor that can be separated
340
Moore and Olsen
Figure 12 Retraction of the mandible and incision of the anterior tonsillar pillar.
Figure 11 Splitting the submental skin and lip and dividing the mylohyoid muscle.
easily under direct vision and included with the specimen (Fig. 15). For deep-lobe tumors that extend through the stylomandibular tunnel into the parapharyngeal space, an initial superficial parotidectomy can be performed. Care should be taken not to compress the tumor against the styloid process, mastoid tip, or pterygoid plates, as this may cause rupture of the tumor and spillage of tumor into the wound. This can lead to increased risk of recurrence. The styloid process or portion of the mastoid tip can be removed to aid in tumor exposure and removal. For malignant tumors of the parapharyngeal space, a total parotidectomy is performed both for complete tumor removal and for adequacy of lymphadenectomy. The superficial and deep portions of the parotid gland both contain lymph nodes that may be involved by malignant salivary tumors of metastatic tumors to the parapharyngeal nodes. Removal of poststyloid tumors of the parapharyngeal space is performed through similar exposure, but removal of the parotid gland is not usually necessary. Only after adequate exposure of the cranial nerves and vessels above and below the tumor should the surgeon begin with tumor removal. The surgeon should get exposure and vessel security between the tumor and the skull base if possible. The uninvolved cranial nerves are isolated from the tumor completely for preservation. After isolation of the cranial nerves, the great vessels are freed from the tumor. For carotid body tumors, the dissection proceeds in the subadventitial plane of
the common carotid artery. The external carotid artery may be preserved, but in many cases division of the external carotid artery aids in mobilization of the tumor and isolation of the common and internal carotid arteries. Dissection proceeds with a combination of bipolar cautery and sharp and blunt dissection until the tumor is totally mobilized and removed.
Figure 13 Exposure of the tumor by mandibulotomy.
Chapter 21: Tumors of the Parapharyngeal Space
341
Figure 15 Photograph of deep-lobe pleomorphic adenoma. Note cuff of normal parotid gland adjacent to tumor.
Figure 14 Exposure of the tumor through cervical–parotid approach and preparation for removal.
In the case of neurilemmomas, the tumor may be able to be dissected away from a portion of the involved nerve. If the nerve can be preserved, then functional recovery may occur. For paragangliomas and neurofibromas, complete resection of the tumor requires division of the nerve above and below the tumor and preservation of function is not possible except with the possibility of nerve grafting. Tumor margins should be checked by frozen section at the proximal and distal ends of division.
Reconstruction Reconstruction is not necessary in most patients with parapharyngeal tumors. After complete removal of the parotid gland, the patient is left with a significant cosmetic defect. If the operation involved resection of the facial nerve or adjacent structures, then the patient may be left with a significant cosmetic or functional defect. Reconstruction of these defects is dependent on patient desires, age, additional treatment plans, expected outcomes, and many other factors. In short, it is highly individualized to the patient. The preoperative discussion should have explored these options in depth to fully guide the surgeon at the time of the operation. We often divide the resection and the reconstruction into separate but intimately coordinated teams. Among the many benefits of this approach is the concept that the resecting surgeon is not influenced subconsciously to minimize the operation because of concerns with the feasibility of reconstructing the defect.
Options for reconstruction include primary closure; dermal fat grafting; muscle transposition with locoregional flaps of the sternocleidomastoid or pectoralis muscles; and microvascular cutaneous, musculocutaneous, and innervated muscular flaps. Again, the reconstruction will be guided by the functional and aesthetic goals of the surgeon and patient. Many patients with total parotidectomy defects can be adequately reconstructed with dermal fat grafting. A skin and subcutaneous free graft approximating the dimensions of the removed gland is harvested from the abdomen and deepithelialized. The graft is suture fixated into position in the parotid bed and the donor site is closed primarily. This graft is less reliable in patients who have been previously radiated or who are planned to undergo postoperative radiation. In these patients, a vascularized free tissue transfer of anterolateral thigh or rectus muscle has been more reliable for us. If patients have extensive resection of bone of the mandible or skull base resected, then vascularized tissue is our reconstructive method of choice.
Closure After tumor removal is completed and the margins are confirmed with frozen section analysis, the wound is prepared for closure. Inspection for bleeding points and correction with suture ligation and bipolar cautery is meticulously performed. The wound is irrigated with saline. If mandibulotomy has been performed, then the mandible is fixated internally with the previously drilled and contoured miniplates placed across the stair-step mandibulotomy. The intraoral incisions are closed with absorbable sutures, and a suction drain is placed in the parapharyngeal space. The skin and subcutaneous layers are closed with absorbable sutures. If a tracheostomy tube has been placed, then it is sutured into position to the skin to prevent accidental decannulation.
POSTOPERATIVE CARE Removal of large parapharyngeal tumors by the cervical– parotic approach leaves significant dead space that is prone to hematoma formation and infection. A broad-spectrum antibiotic is used for several days postoperatively. If the intraoral mucosa has been violated, the use of metronidazole is
342
Moore and Olsen
added. It is essential to maintain constant suction drainage to collapse the cavity left by the tumor. The drain is removed when the 24-hour collection value is negligible. Antibiotics are continued until the drain is removed. Patients with removal of high vagal lesions may have significant dysphagia after the operation. The challenges with dysphagia increase with additional injury to cranial nerves IX and XII. A feeding tube and swallowing therapy may be necessary in some patients. The possibility for aspiration is high, and precautions with positioning during feeding and thickening of liquids may be necessary. In our experience, younger patients and those with isolated vagal injuries recover the ability to swallow early. Older patients and patients with multiple cranial neuropathies may have significantly longer trouble with swallowing, or even permanent dysphagia. If a tracheotomy has been placed for airway precautions, it is only removed once the patient can tolerate occlusion of the tracheostomy during sleep and the airway is well visualized on endoscopy.
Table 3 Complications of Surgery in the Parapharyngeal Space Complication
Sequelae
Recurrent tumor Nerve injury: CN VII, IX, X, and XII Greater auricular nerve Inferior alveolar nerve Cervical sympathetic chain
Additional treatment/death Aspiration Dysphonia Dysarthria Velopharyngeal insufficiency Horner Syndrome Facial paralysis Facial numbness Shoulder dysfunction Hemorrhage Stroke Death Emergency tracheotomy/Death Tooth loss Malunion Nonunion Hypertensive crisis Freys syndrome Soft tissue depression Salivary fistula “First-bite pain” Rhinorrhea Wound drainage Meningitis Delayed wound healing Abscess Fistula Seroma Scar Thrombophlebitis Pneumonia Pulmonary embolism Myocardial infarction
Vessel injury: carotid artery, jugular vein Airway compromise Mandibular osteotomy
Catecholamine-secreting tumor Parotid gland complications
COMPLICATIONS AND AVOIDANCE Complications of surgical approaches in the parapharyngeal space have been reported in a few large series in the literature (18,19).Surgical complications can result in expected sequelae from resection of cranial nerves as a planned part of tumor resection. This can occur frequently in operations of parapharyngeal space masses and can result in both temporary and permanent morbidity. The most frequently reported unexpected complications of surgery in the parapharyngeal space include tumor spillage (20%), first bite pain (12%), trismus and law pain (4%), temporary facial paresis (4%), palatal weakness (2.7%), and cerebrospinal fluid leak (2.0%) (19). Wound infection, orocutaneous fistula, hemorrhage, compromised airway, surgical incision morbidity, and complications of catecholamine-secreting tumors are rare. Fortunately, most operations in the parapharyngeal space are for primary benign neoplasms and have low expected morbidity. In contrast, cancers that extend into the parapharyngeal space or into the infratemporal fossa are associated with significantly higher morbidity after attempts at removal. Reported complications have included major vessel injury, stroke, and death. Table 3 lists the reported complications and sequelae of surgery in the parapharyngeal space (20). Avoidance of complications involves two important steps: (1) adequate and accurate preoperative assessment, and (2) excellent surgical exposure. Accurate preoperative imaging is essential to avoid complications. Both CT and MRI scans help determine the extent of the lesion, the possibility of intracranial extension, and delineate vascular and nonvascular masses. Displacement of the carotid vessels (anteriorly by neurogenic tumors, posteriorly by salivary tumors) can give clues as to origin of the tumor. Angiography can define vascular tumors and give information about carotid blood flow. Fine-needle aspiration can provide helpful information about parapharyngeal tumors and can be performed safely. Cranial nerve monitoring can decrease the incidence of nerve injury in tumors that extend intracranially. Intimate knowledge of anatomy and good meticulous surgical exposure can lead to low parapharyngeal surgical complications. A well-prepared and experienced surgeon can expect to operate in the parapharyngeal space with minimal morbidity and excellent outcomes as long as complications are anticipated and appropriate steps of prevention are taken.
Cerebrospinal fluid leakage
Infection
Other
FOLLOW-UP AND REHABILITATION Timely institution of speech and swallowing rehabilitation can facilitate recovery in patients after resection of neurogenic tumors (21). Laryngoplasty with injectable material or framework medialization can improve dysphonia and aspiration in selected patients with vocal cord paralysis. Unilateral palatal adhesion has been shown to improve the velopharyngeal incompetence and speech in patients with glossopharyngeal nerve sacrifice (22). We usually assess for the need of these procedures several weeks to months after resection to allow for recovery and stability of the patients. In this manner, a more specific and efficient intervention can be planned. The patient should be scheduled for regular followup physical examinations that include nasopharyngeal endoscopy to assess for both the recovery and the possibility of recurrent tumor. Serial imaging may be obtained to help in detection of recurrent tumor. The schedule and type of these examinations is tailored to the location and type of the tumor and the likelihood of regrowth or metastasis. In general, patients with parapharyngeal tumors recover and achieve an excellent quality of life when they have excellent preoperative, operative, and postoperative care. Most patients achieve a high level of functional and cosmetic recovery. For those patients in whom this cannot be expected by the preoperative assessment, then the benefits of surgery should be heavily contemplated by the surgeon and patient before proceeding to the operating room.
Chapter 21: Tumors of the Parapharyngeal Space REFERENCES 1. Work WP, Hybels RL. A study of tumors of the parapharyngeal space. Laryngoscope. 1974;84:1748–1755. 2. Patey DH, Thackray AC. The pathological anatomy and treatment of parotid tumors with retropharyngeal extension (Dumbbell tumors) with a report of 4 personal cases. Br J Surg. 1956– 1957;44:352–358. 3. Robbins KT, Woodson GE. Thyroid carcinoma presenting as a parapharyngeal mass. Head Neck Surg. 1985;7:434–436. 4. Stell PM, Mansfield AO, Stoney PJ. Surgical approaches to Tumors of the Parapharyngeal Space. Am J Otolaryngol. 1985;6:92– 97. 5. Olsen KD. Tumors and Surgery of the Parapharyngeal Space. Laryngoscope. 1994;104:1–28. 6. Carr RJ, Bowerman JE. A review of tumors of the deep lobe of the parotid salivary gland. Br J Oral Maxillofac Surg. 1986;24:155– 168. 7. Warrington G, Emery PJ, Gregory MM. Pleomorphic salivary gland adenomas of the parapharyngeal space: Review of nine cases. J Laryngol Otol. 1981;95:205–218. 8. Chu W, Strawitz JG. Parapharyngeal growth of parotid tumors. Arch Surg. 1977;112:709–711. 9. Green JD, Olsen KD, DeSanto LW. Neoplasms of the vagus nerve. Laryngoscope. 1988;98:648–654. 10. Baysal BE, Ferrell RE, Willett-Brozick JE. Mutations in SDHD, a mitochondrial complex II gene, in hereditary paraganglioma. Science. 2000;287:848–851. 11. Ferlito A, Pesavento G, Recher G. Assessment and treatment of neurogenic and non-neurogenic tumors of the parapharyngeal space. Head Neck Surg. 1984;7:32–43.
343
12. Sobol SM, Wood BG, Conoyer JM. Glossopharyngeal neuralgiaasystole. Head Neck Surg. 1982;90:16–19. 13. Myers En, Johnson JT, Curtin HD. Tumors of the parapharyngeal space in Cancer of the Head and Neck, 4th ed. Saunders, 2003. 14. Whyte AM, Hourhihan MD. The diagnosis of tumors involving the parapharyngeal space by computed tomography. Br J Radiol. 1989;62:526–531. 15. van der May AG, Frijns JH, Cornelisse CJ. Does intervention improve the natural course of glomus tumors: A series of 106 patients seen in a 32 year period. Ann Otolo Rhinol Laryngol. 1992;101:635–642. 16. Evenson LJ, Mendenhall WM, Parsons JT, et al. Radiotherapy in the management of chemodectomas of the carotid body and glomus vagale. Head Neck. 1998;20:609–613. 17. Ducic Y, Oxford L, Pontius AT, et al. Transoral approach to the superomedial parapharyngeal space. Otolaryngol Head Neck Surg. 2006;134:466–470. 18. Carrau RL, Myers EN, Johnson JT, et al. Management of tumors arising in the parapharyngeal space. Laryngoscope. 1990;100:583–589. 19. Hughes KV III, Olsen KO, McCaffrey TV, et al. Parapharyngeal space neoplasms. Head Neck. 1995;17(2):124–130. 20. Moore EJ, Olsen KD. Complications of surgery of the parapharyngeal space. In: Eisele DW, ed. Complications of Head and Neck Surgery. 2nd edition. Mosby, 2008. 21. Netterville JL, Livantos FJ, et al. Rehabilitation of cranial nerve deficits after skull base surgery. Laryngoscope. 1993;102:45– 54. 22. Netterville JL, Vrabec JT. Palatal adhesions. Arch Otolaryngol Head Neck Surg. 1994;120:218–221.
22 Tumors of the Temporal Bone Sam J. Marzo and John P. Leonetti
SURGICAL ANATOMY
ral bone contains bony, cartilaginous, epithelial, respiratory, vascular, paraganglionic, and neural tissues, virtually every type of tumor can occur. Tumors can be benign or malignant, can arise from tissues within or adjacent to the temporal bone, and can also be metastatic to the temporal bone. An understanding of the regional tissue histology can help one develop an effective differential diagnosis when treating patients with temporal bone neoplasms. The EAC, TM, and pinna are covered by squamous epithelium, which can give rise to squamous cell carcinoma (SCCA). Neoplastic change within epidermal elements can also result in basal cell carcinoma and melanoma. The middle ear mucosa is lined by a nonstratified columnar respiratory epithelium. Neoplastic change of this tissue is fortunately rare, but can result in adenomatous tumors of the middle ear. Glomus tumors, which are the most common benign tumors of the temporal bone, arise from paraganglia along Jacobsen nerve (glomus tympanicum tumors) and along the jugular bulb (glomus jugulare tumors). Hemangiomas and schwannomas involving the facial nerve can also rarely occur. In children, the most common malignant tumor of the temporal bone is a rhabdomyosarcoma. Primary tumors of the endolymphatic sac are possible but are fortunately rare. Chondrosarcomas can arise from cartilaginous elements of the lateral skull base. Meningiomas of the posterior fossa and middle fossa can involve the temporal bone. Advanced or recurrent benign and malignant tumors of the parotid gland can involve the temporal bone. Finally, metastatic tumors (breast, prostate, lung, and kidney) are possible, as well as recurrent head and neck carcinoma.
A proper understanding of the surgical anatomy of the temporal bone is critical to successful treatment and management of lesions in this area. The temporal bone has five divisions: mastoid (posterior), squamous (superior), petrous (medial), tympanic (lateral), and zygomatic (anterior). Each of these anatomical areas contains critical structures and is adjacent to important areas of the lateral skull base. Figure 1 is a sagittal section of the human temporal bone highlighting the relationships between the ossicular chain, facial nerve, cochlea, and carotid artery. The middle cranial fossa is superior to the temporal bone; the sigmoid sinus and posterior cranial fossa are posterior and medial; the mandible, glenoid fossa, and parotid gland are anterior; and the cartilaginous pinna with overlying squamous epithelium is lateral. Tumors of the temporal bone can often be relatively asymptomatic. Patients can present with advanced disease. Also, the temporal bone has several inherent weaknesses that allow for spread of disease. Many tumors begin in the external auditory canal (EAC). The EAC is lined by a thin layer of periosteum and squamous epithelium. The medial border of the EAC is the tympanic membrane (TM), which has an outer squamous layer, a middle fibrous layer, and a medial layer of columnar epithelium. Once the TM is breached, the tumor can spread medially along the petrous apex into the carotid canal and petrous carotid artery. Posterior extension into the mastoid can lead to facial nerve invasion and paralysis. Anterior extension through the tympanic bone can result in glenoid fossa/temporomandibular joint involvement and then invasion of the infratemporal fossa. Superior extension through the squamous portion of the temporal bone over the middle ear (called the tegmen) can result in invasion of the dura and middle cranial fossa. A critical understanding of the course of the intratemporal facial nerve is very important for the management of lesions in this area. Figure 2 shows an axial section of the facial nerve as it courses through the temporal bone. The facial nerve has several divisions within the temporal bone. The meatal segment is within the internal auditory canal. Next is the labyrinthine (or geniculate) segment, which bends around the cochlea. The tympanic segment traverses between the horizontal semicircular canal and the stapes. At the pyramidal process (through which courses the stapes tendon), the nerve enters the descending mastoid segment. Next, the nerve leaves the temporal bone through the stylomastoid foramen and enters the parotid gland.
CLINICAL ASSESSMENT Clinical Presentation Symptoms and signs of tumors of the temporal bone are varied and depend on the location of the tumor. Table 2 lists the various clinical presentations (2). Tumors originating in the ear canal can present with a nonhealing lesion, otorrhea, otalgia, or a conductive hearing loss. The most common malignant neoplasm in this location is SCCA. Tumors originating in the middle ear are relatively rare. Glomus tumors are the most common lesions in this area, and can present with pulsatile tinnitus and conductive hearing loss. Adenomas can also present with conductive hearing loss. Facial neuromas are more likely to present with conductive hearing loss than facial paresis. Rarely is there facial twitching or a slow facial paresis. Tumors originating in the internal auditory canal or cerebellopontine angle are likely to present with unilateral tinnitus and or sensorineural hearing loss. The most common tumor in this area is the vestibular schwannoma (or
REGIONAL PATHOLOGY AND DIFFERENTIAL DIAGNOSIS The complete differential diagnosis of temporal bone neoplasms is extensive and is listed in Table 1 (1). As the tempo345
346
Marzo and Leonetti Table 1
Differential Diagnosis of Temporal Bone Neoplasms
External auditory canal Benign
Pleomorphic adenoma Ceruminous adenoma Squamous cell carcinoma Basal cell carcinoma Sarcoma Melanoma Lymphoma Adenocarcinoma Adenoid cystic carcinoma Ceruminous adenocarcinoma
Malignant
Middle ear and mastoid Epithelial neoplasms Benign Figure 1 Sagittal section of human temporal bone showing relationships of facial nerve, cochlea, ossicular chain, and carotid artery.
acoustic neuroma). The tumor begins on the superior or inferior vestibular nerve. Imbalance or vague dizziness is more common than episodic vertigo. Endolymphatic sac tumors can present with similar symptoms. Tumors neighboring the temporal bone can also cause otological symptoms (3). Lesions of the nasopharynx and infratemporal fossa can cause a middle ear effusion from involvement of the eustachian tube. Tumors of the parotid gland involving the temporal bone can present with a fixed mass within the parotid gland, occasionally with facial paralysis. Tumors of the deep lobe of the parotid and infratemporal fossa involving the temporal bone can present with referred otalgia and can mimic trigeminal neuralgia. Hoarseness and dysphagia can signify involvement of the jugular foramen and adjacent cranial nerves.
CLINICAL ASSESSMENT Evaluating patients with tumors of the temporal bone requires a thorough assessment of the presenting symptoms, duration, and prior treatment. It is also important to assess if
Mucosal adenoma Papillary adenoma Papilloma Squamous cell carcinoma Basal cell carcinoma Adenocarcinoma Carcinoid tumor
Malignant
Soft tissue neoplasms Benign
Paraganglioma Myxoma Lipoma Hemangioma Schwannoma Neurofibroma Rhabdomyosarcoma Malignant paraganglioma Malignant schwannoma Hemangiopericytoma
Malignant
Tumors of bone and cartilage Benign
Malignant Miscellaneous Benign
Osteoma Chondroblastoma Giant cell tumor Osteosarcoma Chondrosarcoma Meningioma Teratoma Malignant germ cell tumor
Malignant Malignant lymphomas Metastatic neoplasms
Table 2 Clinical Symptoms of Temporal Bone Neoplasms Symptoms
Signs
Figure 2 Axial section of human temporal bone showing course of intratemporal facial nerve.
Otorrhea Pain Facial weakness Facial numbness Hearing loss Vertigo, imbalance Trismus Hoarseness Dysphagia EAC mass Bloody otorrhea Facial paralysis Other cranial nerve deficits Parotid mass Cervical mass Middle ear mass Middle ear effusion Conductive hearing loss Sensorineural hearing loss
Chapter 22: Tumors of the Temporal Bone
there has been any prior surgery and/or radiotherapy, which can potentially impact the current treatment and necessitate additional reconstructive options after surgery. Other important aspects of the medical history include assessing comorbidities such as hypertension, cardiac disease, cerebrovascular disease, diabetes, chronic pulmonary disease, liver disease, and obstructive sleep apnea. Once a complete history is taken, a complete otolaryngological examination is performed including microscopic otoscopy and cranial nerve testing. Special attention is directed to the facial and lower cranial nerve (4–9) examination. Flexible examination of the larynx is sometimes necessary to accurately assess the vagal nerve. Because the proposed surgery may result in an additional partial or complete unilateral hearing deficit, binaural audiometric testing is required. Some patients will be referred with a tissue diagnosis. Outside pathology should be reviewed for accuracy by the treating institution’s pathology department. In some cases, pathology will not be known before the initial consultation. Granulation tissue in the ear canal (suggesting SCCA) can be biopsied in the office or under general anesthesia if inconclusive. Vascular lesions of the middle ear such as glomus tumors should not be biopsied preoperatively. Palpable parotid and/or neck masses may be amenable to fine-needle aspiration. Finally, for some lesions inaccessible on the physical examination, the diagnosis will rely on imaging studies.
DIAGNOSTIC IMAGING Several key studies are required for management of temporal bone neoplasms. A computed tomographic (CT) scan of the temporal bone will delineate the bony confines of the lesion. It is important to assess for extension of the tumor posteriorly into the mastoid and medially through the TM. The carotid canal and jugular bulb should also be assessed on the CT scan. Figure 3 shows an SCCA of the right temporal bone with extensive bony erosion. MRI is best at showing the soft tissue extent of the lesion. An MRI of the parotid gland with contrast will identify anterior soft tissue extension into the parotid gland and glenoid fossa, inferior extension into the infratemporal fossa,
347
Figure 4 Coronal MRI of lesion of right skull base with intracranial extension.
superior extension into the middle fossa, and posterior extension into the posterior fossa. Figure 4 shows superior extension of an SCCA of the temporal bone into the middle cranial fossa. Further imaging is dictated by the clinical and radiographic assessment. Glomus tympanicum tumors are, by definition, limited to the middle ear and generally do not require angiography. Since many glomus jugulare tumors will require resection of the ipsilateral sigmoid sinus, jugular bulb, and internal jugular vein, angiography can be helpful in delineating arterial feeding vessels and venous outflow patterns.
Therapeutic Imaging For glomus jugulare tumors, a preoperative embolization can be helpful in decreasing intraoperative blood loss. A recurrent or advanced glomus tumor involving the petrous carotid artery in young patient may necessitate preoperative sacrifice of the carotid artery within the petrous portion of the temporal bone with detachable coils, provided the patient has passed a preoperative balloon occlusion test with cerebral blood flow studies (10,11).
Other Imaging Studies Other imaging studies are dictated by the clinical assessment. Patients with presumed metastatic lesions should undergo a metastatic survey, which may include CT scan imaging of the chest and abdomen, and positron emission tomography.
PREOPERATIVE PREPARATION AND ASSESSMENT
Figure 3 bone.
Axial CT scan showing an erosive SCCA of the right temporal
After the clinical and radiographic assessment is performed, the next step is to formulate a treatment plan. Patients with malignant lesions of the temporal bone should be presented at a multidisciplinary tumor board conference. Many patients with malignant lesions of the temporal bone will require adjuvant radiation therapy and possibly adjuvant or neo-adjuvant chemotherapy. The clinical and radiographic assessments of the patient will allow the surgeon to decide if the lesion is technically resectable. The harder question is whether the surgery should be undertaken in every case. In certain circumstances, surgery may not be advisable.
348
Marzo and Leonetti
Relative contraindications for surgical resection include malignant lesions involving the cavernous sinus, carotid artery, infratemporal fossa, and paraspinous musculature (12). However, some have reported extended survival even with intracranial extension (4). Some of these patients will benefit from palliative chemoradiotherapy. Also important in the preoperative assessment of patients is to assess reconstructive options. Can the defect be closed primarily by oversewing the ear canal and obliterating the mastoid cavity with abdominal fat? Will a local regional flap (temporalis) or rotational flap (trapezius or latissimus dorsi) be required, or is a microvascular tissue transfer needed? Patients who have undergone prior radiotherapy will benefit from a vascularized tissue closure to decrease postoperative complications. Involving a reconstructive surgeon in the patient’s care preoperatively is essential in these cases. For lesions involving or abutting the middle fossa or posterior fossa, a preoperative neurosurgical consultation is important. Finally, patients with significant comorbidities can benefit from a preoperative medical, pulmonary, or cardiac consultation. Finally, as many of these surgical cases can last several hours, patients with poor pulmonary reserve should be offered a temporary tracheotomy, which can reduce postoperative pulmonary complications. Once the treatment plan has been formulated, a detailed discussion with the patient and supportive family members is important. The goals, risks, potential complications, and expected postoperative course should be discussed. Also if cranial nerve deficits are anticipated, a treatment plan to address these deficits is also discussed preoperatively. If multiple lower cranial nerve deficits are possible, the discussion should also mention the possibility of tracheotomy and feeding tube placement for airway protection and nutrition, respectively. If postoperative radiotherapy is likely to be needed, the patient should undergo a preoperative consultation with radiation oncology. As treatment of patients with such tumors is a multidisciplinary effort, communication and coordination between disciplines is essential. This best occurs through the development of skull base teams involving head and neck oncologists, neurotologists, neuroanesthesiologists, interventional neuroradiologists, radiation oncologists, plastic and reconstructive surgeons, surgical assistants, and intraoperative cranial nerve monitoring personnel.
SURGICAL TECHNIQUE General Principles The clinical assessment, pathology review, and radiographic assessment will dictate the extent of surgery. Important considerations include the status of the ear canal, middle ear and conductive hearing apparatus, cochlea and vestibular apparatus, and potential for cerebrospinal fluid otorrhea or rhinorrhea. For example, if an SCCA of the ear canal is treated with a sleeve resection of the ear canal, skin grafting, and postoperative radiotherapy, the patient may likely develop a chronic problematic mastoid cavity that has to be debrided at regular intervals, heals poorly, periodically drains or gets infected, and is prone to serous middle ear effusions. This patient may have been better served with a lateral temporal bone resection with oversewing of the ear canal, obliteration of the mastoid cavity with abdominal fat, and hearing rehabilitation with a bone-anchored hearing aid (5). The best way to address potential postoperative problems with CSF is to preoperatively include plans for repair.
CSF wound accumulations and fistulae can be problematic and eventually lead to meningitis. Therefore, incisions should be multilayered. Abdominal fat can be used to obliterate dead space. Finally, the eustachian tube and middle ear may need to be obliterated with muscle to prevent postoperative CSF rhinorrhea. The location of tumors within and adjacent to the temporal bone determines the surgical approach. Benign tumors limited to the middle ear and mastoid can be treated with a tympanomastoidectomy. In general, tumors limited to the ear canal can be treated with a temporal bone resection. The ear canal is oversewn and the cavity is packed with abdominal fat. Depending on the stage of the tumor, a parotidectomy and neck dissection may also be required. If the pinna is involved and has to be resected, the surgical procedure is as above; however, a rotational flap (trapezius) or free flap reconstruction (rectus abdominus) is generally required. Tumors of the jugular foramen are medial to the descending mastoid segment of the facial nerve and therefore may require that the nerve be mobilized anteriorly. The ear canal limits the potential mobilization. For extensive glomus jugulare tumors involving the petrous carotid artery, the ear canal is removed and the facial nerve is mobilized at the tympanic segment. Craniotomies may be necessary for temporal bone tumors with posterior or superior extension.
Patient Positioning Preoperative consultation with the anesthesiologist is critical to ensure that the short-acting paralytics are used for the anesthetic induction so that cranial nerve monitoring will be possible throughout the case. If a tracheotomy is planned, it is generally done after induction of anesthesia. The operative table is then turned 180 degrees so that the scrub nurse is positioned opposite the surgeon. Most patients will be positioned supine, with the treated ear up and the head gently turned to the opposite side. Overtorsion of the head can occlude the contralateral jugular venous system, which can cause cerebral edema if the ipsilateral venous system will be taken surgically. The head can be immobilized with tape, placed in a Mayfield head holder, or even secured with pins (if a craniotomy is anticipated). If an abdominal free flap is anticipated for wound closure, it can be harvested with the patient supine. If the reconstructive option is a pedicled flap such as a trapezius or latissimus, the patient can be positioned on his or her side for both the surgical resection and reconstruction. All bony prominences are padded. Arterial lines are placed. A foley catheter is placed for monitoring urine output. Cranial nerve monitoring electrodes are also placed. If a rectus abdominus free-tissue transfer or fat graft is anticipated for wound closure, the abdomen is also prepared accordingly. Figure 5 shows patient positioning and room set up for a lateral temporal bone resection.
Surgical Incision The best way to prevent postoperative wound healing issues and wound complications is by accurate planning of the surgical incisions and approach. Most tumors of the temporal bone can be approached through a C-shaped postauricular incision. The incision can be modified as needed. It is placed further behind the ear if more access is required posteriorly, along the sigmoid sinus. If superior extension is required (such as for a middle fossa craniotomy), the incision is placed more superiorly, within the temporal hairline. If access to the neck and great vessels is required, the incision gently curves into a cervical neck crease. If the ear canal is to be part of
Chapter 22: Tumors of the Temporal Bone
349
Soft Tissue Dissection
Figure 5 Axial section of a right temporal bone showing relationships between ear canal facial nerve and middle and inner ear.
the surgical resection, an incision is made in the cartilaginous EAC encompassing the lesion (Fig. 6). The incision is made initially with a 10-bladed scalpel. Undermining of the skin above the temporalis fascia is then quickly performed for approximately 1 cm. Manual pressure on the wound edges will temporarily tamponade arterial bleeding. Hemostatic clips are then placed. These are preferred to electrocautery of the skin edges as the former will allow for improved vascularity of the skin incision and thus quicker wound healing. This is especially important if the middle or posterior fossa dura needs to be opened as part of the surgical procedure and CSF can enter the wound.
The soft tissue dissection should be planned to incorporate stepped incisions and multiple layers to allow watertight closure, especially if CSF can enter the wound. Next, a scalpel, dissection scissors, or electrocautery is used to develop the plane above the temporalis fascia. A stepped incision is made through the fascia, muscle, and periosteum. A C-shaped incision 1 cm inside the surgical incision is easier to close in a watertight fashion than a T-shaped incision, which is commonly used in otological surgery. This flap is elevated anteriorly until the root of the zygoma, soft tissue entering the ear canal, and mastoid tip are identified. Elevation of the soft tissues overlying the mastoid cortex should be done in the subperiosteal plane. The soft tissues are reflected anteriorly, out of the surgical field, moistened with gauze-soaked saline, and secured with hooks. Access to the neck contents and great vasculature is possible through a cervical extension of the incision. If ear canal sacrifice is required, it is sectioned at the bony cartilaginous junction, and closed in multiple layers. Carrying this dissection forward above the parotid fascia will allow further anterior extension, especially if a parotidectomy is required. Modifications of the soft tissue dissection are also possible depending on the anticipated size of the surgical defect. The temporoparietal fascia can be preserved and reflected forward as a rotational flap reconstruction. The temporalis muscle can also be used as a rotational flap, based on its deep vascular supply.
Bony Dissection and Osteotomies Once the soft tissue dissection is completed, the bony dissection begins with a large cutting burr and copious irrigation. Most approaches to temporal bone neoplasms will require some form of mastoidectomy. Principles of safe bony
Figure 6 Patient positioning for temporal bone surgery.
350
Marzo and Leonetti Table 3 Anatomical Location and Surgical Approach for Paraganglioma Tumor Glomus tympanicum Limited to middle ear Extension into mastoid Glomus jugulare Limited to jugular bulb Involving petrous carotid Intracranial extension
Figure 7 Outline of skin incisions for a lateral temporal bone resection including resection of skin of the EAC.
dissection include wide saucerization and the utilization of constant landmarks for finding key structures. The deepest portion of the dissection should always be anterosuperior, as the facial nerve and horizontal semicircular canal are deepest here. Superiorly, the tegmen or bony plate overlying the middle cranial fossa dura is carefully skeletonized. This plate usually has an irregular shape (like that of a potato chip) rather than a flat plane, so care is taken to avoid dural injury. Anteriorly, the bony ear canal, which is cylindrically shaped, is also carefully skeletonized. Posteriorly, the sigmoid sinus is identified and skeletonized. Finally, inferiorly the mastoid tip is identified. Figure 7 shows the relationships of the mastoid, middle ear, facial nerve, and ear canal within the temporal bone. As the dissection proceeds medially, the drill size and suction irrigator size are decreased accordingly, as the size of the surgical field narrows. Also, while the surgeon can perform a simple mastoidectomy without magnification, using an operating microscope will allow safe dissection of the intratemporal facial nerve, middle ear, and inner ear structures. Variations in the basic mastoidectomy are performed depending on the disease process, its clinical presentation, and radiographic imaging and staging. For paragangliomas of the temporal bone, Table 3 lists the location and corresponding surgical approach. Small glomus tympanic tumors limited to the middle ear can be approached through a postauricular/transcanal approach (Fig. 8). The TM is reflected anteriorly and the tumor is carefully removed as it is coagulated with bipolar cautery (Video 1). Tumors limited to the jugular bulb can be resected with an intact canal wall transmastoid and transcervical approach. As the jugular bulb is medial to the mastoid portion of the facial nerve and stylomastoid foramen, it may be necessary to mobilize this portion of the nerve anteriorly. This mobilization is limited by the bony EAC. VIDEO 1 Intraoperative video of a glomus tympanicum tumor seen through a postauricular/transcanal approach. The tumor is coagulated with bipolar cautery before removal.
Surgical approach Transcanal Extended facial recess Transmastoid/cervical with possible facial rerouting Infratemporal fossa Infratemporal fossa with intracranial extension
For larger jugulotympanic paragangliomas with extension along the petrous carotid artery, it is necessary to perform an infratemporal fossa approach, as described by Fisch (6). This approach involves a wide mastoidectomy, as described above, resection of the tympanic bone and soft tissue of the ear canal with oversewing of the soft tissue of the external auditory meatus. The facial nerve is skeletonized within the tympanic and mastoid segments and mobilized anteriorly. Using continuous facial nerve monitoring and preserving the periosteum of the stylomastoid foramen during this portion of the rerouting can decrease incidence of postoperative facial nerve paresis (7). The internal jugular vein, internal carotid artery and lower cranial nerves are identified in the neck. The internal jugular vein is divided. The sigmoid sinus is occluded. The adventitia of the lateral portion of the jugular bulb is removed with the tumor, with attempted preservation of the lower cranial nerves and internal carotid artery. Bleeding from the inferior petrosal sinus and condylar vein can be quite brisk and is controlled with oxidized cellulose. Intracranial extension of the tumor is resected in conjunction with neurosurgery after the temporal portion of the tumor is removed. Closure includes obliteration of the eustachian tube with muscle, waxing of open mastoid air cells, temporalis fascia grafting of any dural defect, and obliteration of the cavity with abdominal fat. A pressure dressing is also placed for several days postoperatively. Postoperative care is described in the next section.
Figure 8 Intraoperative photograph of a glomus tympanicum tumor.
Chapter 22: Tumors of the Temporal Bone
351
Table 4 SCCA Stage and Surgical Approach Stage
Definition
Surgical approach
T1 T2 T3
Limited to EAC without bony erosion or evidence of soft tissue extension Limited bony EAC erosion or limited ≤5 mm soft tissue involvement Full thickness EAC erosion with limited ≤5 mm soft tissue involvement or middle ear involvement, or facial paresis Erosion of cochlea, petrous apex, medial wall of middle ear, carotid canal, jugular foramen dura, and >5mm soft tissue involvement
Lateral temporal bone resection Lateral temporal bone resection Subtemporal temporal bone resection
T4
Total temporal bone resection
While benign tumors of the temporal bone can be resected in a piecemeal approach, malignant tumors, such as SCCA, or advanced parotid malignancies require an en bloc approach. Table 4 lists the clinical stage and corresponding surgical approach for SCCA of the temporal bone (8). For T1 and T2, lesions, a lateral temporal bone resection is performed (9). A negative surgical margin is obtained at the ear canal and the canal is oversewn. A partial or total auriculectomy may be necessary depending on tumor extension. An intact canal wall tympanomastoidectomy is performed with an extended facial recess approach. Dissection continues around the bony ear canal into the glenoid fossa anteriorly. The incudostapedial joint is sectioned and the incus is removed. The tensor tympani is sectioned. The stylomastoid foramen is widely exposed and decompressed. The digastric muscle is exposed in the mastoid tip. A chisel is used to section the tympanic bone. The specimen is then reflected anteriorly. Parotidectomy and supraomohyoid neck dissections are performed for staging purposes (Fig. 9). Closure is as described for the infratemporal fossa approach. A trapezius rotational flap or rectus flap may be required if an auriculectomy has also been performed.
Most patients will also undergo postoperative radiotherapy, and the 5-year survival for tumor limited to the EAC is approximately 50% to 70% (13,14). Once SCCA breaches the ear canal and enters the mastoid and middle ear, it can spread easily within the temporal bone. For tumors limited to the middle ear and mastoid, a subtotal temporal bone resection is performed, which removes the temporal bone lateral to the petrous apex, including the otic capsule. This operation includes mobilization of the petrous carotid artery as well as an infratemporal fossa dissection (Fig. 10). Five-year survival for patients treated with subtotal temporal bone resection ranges from 35% to 41.7% (13,14). Once the tumor invades the petrous apex, it can invade the dura, brain parenchyma, and internal carotid artery. The only hope at cure is a total temporal bone resection, which includes resection of the petrous apex with or without resection of the petrous carotid artery (15). This approach involves an extensive infratemporal fossa dissection with carotid artery mobilization and/or resection and middle and posterior fossa craniotomies. Five-year survival is poor and ranges from 0%
Figure 9 Schematic drawing of a lateral temporal bone resection. The specimen has been reflected anteriorly and includes a lateral parotidectomy.
Figure 10 Schematic drawing of subtotal temporal bone resection. The petrous carotid artery has been preserved and mobilized inferiorly. An osteotome is placed in the petrous canal and angled toward the internal auditory canal.
352
Marzo and Leonetti
to 11% (13). Some authors believe that invasions of the cavernous sinus, carotid artery, paraspinal musculature, and infratemporal fossa are contraindications to surgical resection (12). Other authors have reported extended survival with dural and intracranial extension treated with surgery and postoperative radiotherapy (4). Preoperative chemoradiotherapy combined with surgical resection may have a role in some of these patients (16). Closure is as above.
may require continuous lumbar drainage. As humans produce about 18 cc of CSF per hour, less than this is removed, typically 10 cc/ hr. Removing CSF too rapidly can lead to tension pneumocephalus, mental status changes, and even brain herniation (17,18). CSF rhinorrhea should be treated with reoperation and packing of the middle ear, eustachian tube, and opened mastoid air cells.
FOLLOW-UP AND REHABILITATION POSTOPERATIVE CARE The care varies with the extent of the surgery. Patients with glomus tympanicum tumors can undergo surgery in an outpatient surgery center. If the facial nerve has been mobilized or extensive soft tissue dissection has been performed, the patient can be observed overnight and discharged the following day. If there has been jugular venous system dissection, carotid artery dissection, or transdural surgery, the patient should be observed in the neurological intensive care unit and watched closely hemodynamically and neurologically. Any change in the patients’ mental status should be addressed immediately, as this may suggest intracranial hemorrhage and/or meningitis. Postoperative cranial nerve deficits are addressed as described below. Discharge planning includes education of the patient and caretakers as to what to watch for regarding the wound and other issues. Redness, purulent drainage, clear drainage, fluid accumulations under the wound, changes in mental status, fever, difficulty in breathing, and calf pain all merit evaluation in the office and appropriate intervention. CSF wound accumulations can be managed with repeated sterile aspiration and pressure dressings. The patient is followed closely until the wound is completely healed. Stitches and staples are removed 1 to 2 weeks after surgery.
COMPLICATIONS AND AVOIDANCE Any surgical team that treats patients with temporal bone neoplasms will encounter complications. One should continually strive to decrease the complication rate. The first place to start is to spend time preoperatively making sure the patient is adequately prepared for surgery from a medical standpoint and that the patient and family members understand the goals of surgery and potential complications. The surgeon should also thoroughly review the imaging studies to make sure the proposed surgical approach will be ideal. Communication between all members of the surgical team is essential. Preoperative embolization of vascular temporal bone neoplasms can decrease intraoperative blood loss. Intraoperatively, the patient is carefully positioned to avoid potential cervical spine and brachial plexus injury. Cranial nerve monitoring can facilitate cranial nerve preservation. The patient is kept stable hemodynamically during the procedure. Care is taken around the dural venous sinuses. Overpacking the sigmoid sinus can lead to obstruction of the vein of Labbe and venous infarction of the temporal lobe. Excessive packing of the inferior petrosal sinus can injure the 9th, 10th, and 11th cranial nerves. Judicious use of mannitol during the surgical procedure can decrease the volume of the intracranial space in patients requiring transdural procedures. Postoperative cerebrospinal fluid wound issues may lead to meningitis. Sterile wound CSF accumulations may be treated with repeated sterile aspiration and pressure dressings. CSF wound fistulae may require reapproximation of the wound edges. Persistent fistulae or wound accumulations
Surgery involving the glossopharyngeal and vagal nerve may lead to severe swallowing deficits with the potential for aspiration. These patients should undergo a bedside vocal and swallowing evaluation using flexible fiber-optic laryngoscopy postoperatively before being allowed to take oral feeding (19). Some patients will require adjunctive procedures including thyroplasty (20) and possibly palatal adhesion (21). Feeding tube placement may be necessary in some patients. Ultimately, time and speech therapy will improve postoperative lower cranial nerve function. Rehabilitation of the facial nerve and eye begins immediately postoperatively. A paralytic eyelid risks exposure keratitis and possible blindness. Oculoplastic consultation can be very helpful. For temporary facial paresis, the eye is lubricated and taped closed at night. Long-standing paresis is best treated with a gold weight (22) and possible canthoplasty (23). Lower facial paresis is managed conservatively when the nerve is intact. If the nerve function fails to improve, nerve substitution procedures (7–12 grafting) or static suspension procedures may be required. Return to work status depends on the vigor of the work activity and the patient’s motivation to return to work. Patients working at heights or using heavy machinery in whom a postoperative vestibular or neurological deficit exists may not be able to return to these jobs, but may perform some form of sedentary or office work. Temporary or short-term disability may be necessary in other patients. Postoperative radiotherapy is often required in patients with malignant neoplasms of the temporal bone. This usually begins at about 6 weeks postoperatively, as the wound is usually healed at this point. It is important to follow the patient closely during postoperative radiotherapy as breakdown of the skin of the postauricular area and pinna can occur. This is initially managed conservatively. Exposed cartilage of the pinna may require debridement. Postoperative imaging studies are necessary in some patients to assess for recurrent or residual disease. MRI is usually performed 6 months after the initial procedure and then biannually or annually. Postsurgical changes will generally remain stable over time, while lesions that enhance and enlarge represent residual or recurrent disease. In summary, patients with neoplasms of the temporal bone often require multidisciplinary management. A thorough understanding of the microsurgical anatomy and various pathologies of the temporal bone and neighboring lateral skull base are critical for proper treatment of diseases in this area. Postoperative adjuvant therapy may be required for malignant lesions. Lower cranial nerve deficits will require time and possibly further adjunctive surgical procedures for adequate rehabilitation. REFERENCES 1. Wackym PA, Friedman I. Unusual tumors of the middle ear and mastoid. In: Jackler RK, Driscol CLW, eds. Tumors of the ear and temporal bone. Philadelphia, PA: Lippincott Williams & Wilkings, 2000:128–145.
Chapter 22: Tumors of the Temporal Bone 2. Leonetti JP, Marzo SJ. Malignancy of the temporal bone. Otolaryngol Clin N Am. 2002;35:405–410. 3. Leonetti JP, Smith PG, Kletzker R, et al. Invasion patterns of advanced temporal bone malignancies. Am J Otol. 1996;17:438– 442. 4. Moffat DA, Grey P, Ballagh RH, et al. Extended temporal bone resection for squamous cell carcinoma. Otolaryngol Head Neck Surg. 1997;116:617–623. 5. Tjellstrom A. Osseointegrated systems and their application in the head and neck. Adv Otolaryngol Head Neck Surg. 1989;3:39– 70. 6. Fisch U. Infratemporal fossa approach for glomus tumors of the temporal bone. Ann Otol Rhinol Laryngol. 1982;91:474–479. 7. Leonetti JP, Brackmann DE, Prass RL. Improved preservation of facial nerve function in the infratemporal approach to the skull base. Otolaryngol Head Neck Surg. 1989;101:74. 8. Arriaga M, Curtin H, Hirsch BE, et al. Staging proposal for external auditory meatus carcinoma based on preoperative clinical examination and CT findings. Ann Otol Rhinol Laryngol. 1990;99:714–721. 9. Crabtree JA, Britton BH, Pierce MK. Carcinoma of the external auditory canal. Laryngoscope. 1976;86:405–415. 10. Peterman SB, Taylor A, Hoffman JC. Improved detection of cerebral hypoperfusion with internal carotid balloon test occlusion and 99mTc-HMPAO SPECT imaging. AJNR. 1991;12:1035–1044. 11. Monsein LH, Jeffery PJ, Van Heerden BB, et al. Assessing adequacy of collateral circulation during balloon test occlusion of the internal carotid artery with 99mTc-HMPAO SPECT. AJNR. 1991;12:1045–1051. 12. Pensak ML, Gleich LL, Gluckman JL, et al. Temporal bone carcinoma:Contemporary perspectives in the skull base surgical area.Laryngoscope. 1996;106:1234–1237.
353
13. Prasad S, Janecka IP. Efficacy of surgical treatments for squamous cell carcinoma of the temporal bone: A literature review. Otolaryngol Head Neck Surg. 1994;110:270–280. 14. Lewis JS. Temporal bone resection:review of 100 cases. Arch Otolaryngol. 1975;101:23–25. 15. Graham MD, Sataloff RT, Kemik JL, et al. Total en bloc resection of the temporal bone and carotid artery for malignant tumors of the ear and temporal bone.Laryngoscope. 1984;94:528–533. 16. Nakagawa T, Kumamato Y, Natori Y, et al.Squamous cell carcinoma of the external auditory canal and middle ear: An operation combined with preoperative chemoradiotherapy and a free surgical margin. Otol Neurotol. 2006;27:242–249. 17. Effron MZ, Black FO, Burns D. Tension pneumocephalus complicating the treatment of postoperative CSF otorrhea. Arch Otolaryngol Head Neck Surg. 1981;107:579–580. 18. Graf CJ, Gross CE, Beck DW. Complication of spinal drainage in the management of cerebrospinal fluid fistula:Report of three cases. J Neurosurg. 1981;54:392–395. 19. Bastian RW. Videoendoscopic evaluation of patients with dysphagia: An adjunct to the modified barium swallow. Otolaryngol Head Neck Surg. 1991;104:339–342. 20. Netterville JL, Jackson CG, Civantos FJ. Thyroplasty in the functional rehabilitation of neurotologic skull base surgery patients. Am J Otol. 1993;14:460–465. 21. Netterville JL, Vrabec JT. Unilateral palatal adhesion for paralysis after high vagal injury. Arch Otolaryngol Head Neck Surg. 1994;120:218–224. 22. Jobe RP. A technique for lid-loading in the management of lagophthalmos in facial paralysis. Plast Reconstr Surg. 1974;53:29–31. 23. Tenzel RR. Treatment of lagophthalmos of the lower lid. Arch Ophthalmol. 1969;81:366–368.
23 Evaluation and Management of Sellar Tumors Jay Jagannathan, Edward R. Laws, and John A. Jane, Jr.
flow voids. Within the contrast enhanced cavernous sinus, portions of cranial nerves III, IV, V, and VI are rarely seen on the coronal or axial images. In coronal images, cranial nerve III lies lateral and slightly superior to the carotid artery. The course of cranial nerve VI parallels the internal carotid artery in axial images. Since the anatomy of the sphenoid sinus is critical in the majority of surgical approaches to the sella, this warrants special mention. In spite of the complex anatomy and important surgical relationships of the sphenoid sinus, very few chapters focus on it. The sphenoid sinus is extremely variable with respect to size, shape, and relation to the sella. Oftentimes, it can contain septations, which can complicate its surgical anatomy, and can contain varying degrees of pneumatization, particularly in pediatric patients. The sphenoid sinus has been described as being postsellar, presellar, or conchal. The postsellar type of sphenoid sinus is well pneumatized with bulging of the sellar floor into the sinus. The presellar type of sinus is situated in the anterior sphenoid bone and does not penetrate beyond the perpendicular plate of the tuberculum sellae. The conchal type of sphenoid sinus does not reach into the body of the sphenoid bone, and its anterior wall is separated from the sella turcica by approximately 10 mm of cancellous bone. In cases of extensive pneumatization, the maxillary nerve may bulge into the lateral wall of the sphenoid sinus. In extreme cases, the nerve may be entirely surrounded by pneumatization. Identifying the ostia of the sphenoid sinus is critical for both endoscopic and microscopic approaches to the pituitary (see section on transsphenoidal surgery below). The ostia are usually located in the sphenoethmoidal recess, medial to the superior or supreme turbinate, where they can usually be seen well with the endoscope. The floor of sphenoid sinus is occasionally composed of ridges that cover the vidian nerve. The medial and superior walls are usually smooth and the superior wall may balloon outward from pressure of the sella turcica (hypophysis). Two bulges on the lateral wall of the sinus are of considerable clinical significance. They are produced by the optic nerve and the internal carotid artery. Depending on the degree of pneumatization, these two bulges may be barely noticeable or very obvious. Magnetic resonance (MR) imaging is the most sensitive method of radiographic assessment of the pituitary (6–8), although the sensitivity of this technique in detecting small pituitary lesions is variable, and can be as low as 0.22 cm in patients with Cushing disease (6,7,9). In cases where T1-weighted MR imaging is negative or equivocal, a variety of more sensitive imaging techniques have shown promise in detecting adenomas in secretory tumors (10). An angiogram or MRA may be used in cases where an aneurysm or pseudoaneurysm is suspected, and
INTRODUCTION Sellar and parasellar lesions include a diverse group of tumors. The majority of tumors in this region are pituitary adenomas, although dysembryogenic lesions of the midline, such as Rathke cleft cysts, also occur. Suprasellar lesions include craniopharyngiomas, germinomas, dermoid/epidermoid cysts, meningiomas, lipomas, teratomas, metastases, and hamartomas (1–3). Recent advances in microsurgery, surgical endoscopy, neuroimaging, and molecular biology have changed the diagnosis and management of pituitary tumors (4). In this chapter, we focus on current concepts in the understanding and management of these diverse pathological entities.
SURGICAL ANATOMY OF THE SELLAR REGION AND DIAGNOSTIC IMAGING An understanding of sellar and parasellar anatomy and development is critical to the diagnosis of pathology in this region. The pituitary gland is composed of anterior and posterior lobes that develop separately before fusing during embryogenesis. The anterior lobe (adenohypophysis) arises from Rathke pouch in the primordial stomodeum and then migrates along the craniopharyngeal duct to the sella. The posterior lobe (neurohypophysis) arises from an evagination of the inferior third ventricle and then migrates downward into the sella (5). The sella turcica is a shallow depression in the sphenoid bone bounded anteriorly and inferiorly by the sphenoid sinus, laterally by the cavernous sinuses, posteriorly by the dorsum sellae, and superiorly by the diaphragma sellae and suprasellar cistern. In normal adults, the adenohypophysis appears isointense to cerebral white matter on standard T1-weighted magnetic resonance (MR) images. The neurohypophysis may appear hyperintense to the adenohypophysis on T1-weighted images due to the effect of stored neurosecretory granules on relaxation time. This hyperintense region, the so-called posterior pituitary “bright spot,” is evident in up to 90% of children but is less consistently observed in adults (Fig. 1). Parasellar anatomy is seen best in contrast-enhanced MR images (Fig. 2). The cavernous sinus, which lies directly lateral to the pituitary gland, enhances after the administration of gadolinium-containing contrast media. Its enhancement is normally greater than that of the adjacent pituitary gland so that the border between the gland and the cavernous sinus can be discerned. The internal carotid artery within the cavernous sinus appears as a signal void on MRI images because the protons within flowing blood do not produce detectable signal on T1- or T2-weighted pulse sequences. On sagittal images, the carotid siphon may be seen in profile and in coronal images it is seen in cross section as several round 355
356
Jagannathan et al.
Figure 1 The classic “bright spot” representing the posterior pituitary gland is visualized in this nonenhanced sagittal MR image.
diffusion-weighted imaging may be used in cases where an epidermoid tumor is suspected. Special MR sequences are helpful in diagnosing specific tumor subtypes, particularly secretory adenomas. This is particularly relevant in Cushing disease, where tumors are often undetectable on MR imaging. Spoiled gradient-recalled acquisition in the steady-state (SPGR) MR imaging involves acquiring thin sections of 1 mm thickness (compared with 2.5–3 mm on standard spin echo (SE) sequences), using a spoiler gradient to shorten repetition time (11) [Fig. 3(A)]. The SPGR technique allows for a substantial improvement in the spatial resolution of images that are acquired. In a study of 30 children with Cushing disease, Batista et al (12) reported the overall probability that postcontrast SE-MR imaging would be positive in a child with surgically proven Cushing disease was 25%, while the probability of postcontrasted SPGR-MR imaging on the same patient was 71%. Similar data have been reported in adults as well (11).
Figure 2 Normal sellar anatomy on coronal T1-weighted contrast enhanced MR imaging. The normal pituitary gland (white arrowhead) is bounded by the carotid arteries laterally (red arrows). Superiorly, the optic chiasm can be seen in this cut (white arrow). In between the chiasm and the gland is the pituitary stalk.
Figure 3 (A, left panel) SPGR-MR imaging and (B, right panel) dynamic MR imaging may be used to localize adenomas in cases where standard postcontrast MR imaging is nondiagnostic. (Reproduced with permission) (13).
Dynamic MR imaging takes advantage of a tumor’s characteristic slow constant enhancement when compared to the normal pituitary (14). Studies have demonstrated that the timing of postcontrast imaging is critical in diagnosing an adenoma, since the adenoma–normal pituitary maximal contrast following gadolinium injection may occur in some cases within seconds and last for only a few minutes (15,16). Modern MR technology now allows further dynamic studies, such as acquisition of images within seconds after the injection of gadolinium and allowing dynamic MR imaging to capture the early maximal contrast differences between tumors and normal pituitary tissue [Fig. 3(B)].
EPIDEMIOLOGY Pituitary adenomas are the most common cause of pituitary malfunction in adults. They represent 15% to 20% of surgically treated primary brain tumors, with a low incidence in childhood that increases during adolescence (1,17). Although the incidence varies according to age, sex, and ethnic group, between 0.5 and 10.7 per 100,000 in the population are diagnosed annually with a pituitary adenoma. Autopsy series indicate that pituitary tumors are quite common and that nearly 25% of the population may harbor undiagnosed adenomas (18,19). The majority of these tumors are less than 3 to 5 mm in diameter and would not require medical or surgical intervention. Although women are reported to harbor pituitary tumors more frequently than men do, this may reflect the relative contribution of prolactinomas and adrenocorticotropic hormone (ACTH)-secreting tumors, both of which have a female predominance (20). Among the varying classes of adenomas, prolactinomas and clinically nonfunctioning adenomas have the highest proportions, and account for nearly two-thirds of all pituitary tumors with a peak incidence in women between the ages of 20 and 50. Prolactin (PRL-) secreting adenomas also account for 40% to 60% of functioning adenomas and are the most common subtype of pituitary tumor diagnosed in adolescents (21). Growth hormone (GH)-secreting adenomas represent nearly 30% of all functioning tumors. Nearly three-quarters of GH-secreting adenomas are macroadenomas. Approximately 40 to 60 individuals per million have acromegaly (22,23). Between 3 and 4 new cases per million are diagnosed annually (22–27). Most present in their third to fifth decades after they have been developing symptoms and signs for many
Chapter 23: Evaluation and Management of Sellar Tumors
Figure 4 Sagittal (left panel) and coronal (right panel) T1-weighted contrast enhanced MR imaging, depicting a large craniopharyngioma in an 8year-old boy who presented with growth delay and gradual visual decline. Optic nerve involvement is best visualized on the sagittal image (white arrow).
years (27). Acromegaly is associated with an increased incidence of cardiovascular, respiratory, cerebrovascular, and malignant disease. Accordingly, studies report an increased risk of mortality compared to the unaffected population (27). Although some studies report a higher incidence of several cancers, others have only confirmed an increase risk of colon cancer (28–31). There is some evidence that mortality risk may be different between the sexes, but other reports find similar degrees of increased mortality in both sexes (27,32). Still others report increased risks of death in men from cardiovascular, respiratory, cerebrovascular, and malignant diseases, but primarily from cerebrovascular disease in women (33,34). ACTH adenomas account for 15% to 25% of all functioning adenomas and are the most common pituitary tumors diagnosed in prepubertal children (1,2,17,35). The majority of ACTH adenomas, regardless of age, are microadenomas. Approximately 39 individuals per million have Cushing disease and the annual incidence is estimated at 2.4 per million (18). Cushing disease is more common in women, who tend to present in their third and fourth decades (18). There is a high incidence of hypertension and diabetes mellitus as well as higher vascular disease-related mortality (10,36). Craniopharyngiomas account for the overwhelming majority (approximately 90%) of neoplasms arising in the pituitary region (other than pituitary adenomas) in children, but are rare in adults (3–5% of adult tumors) (37,38). Most of these lesions arise from the Rathke pouch—a cystic diverticulum that originates from between the anterior and posterior pituitary gland. Patients with craniopharyngiomas have a bimodal age distribution during the first and second decade of life and then again in the fifth, and there is no apparent predilection for sex (Fig. 4) (39). Most of them originate in the intrasellar and suprasellar region (70%), with suprasellar localizations (20%) or solely intrasellar lesions (10%) occurring less frequently (40).
CLINICAL PRESENTATION AND DIAGNOSTIC EVALUATION PRIOR TO SURGERY Symptoms and clinical signs of tumors in the region of the sella turcica depend upon the type and size of the tumor and age of the patient. The patient may be asymptomatic and the lesion may be discovered during imaging for an unrelated condition. Mass effect from a sellar tumor may produce variable endocrinological and neurological manifestations. Di-
357
Figure 5 In contrast to the craniopharyngioma in Figure 3, this 62-yearold man presented with sudden visual decline and hypopituitarism. (Left panel) Nonenhanced coronal MR imaging demonstrated a large nonfunctioning tumor with varying signal intensity, consistent with blood. (Right panel) Following emergent transsphenoidal resection, the patient regained vision, and no residual tumor could be seen on follow-up MR imaging.
minished growth velocity or short stature is a common feature in many children harboring pituitary adenomas; this may be accompanied by delayed puberty or hypogonadism. Mass effect can also produce galactorrhea (from hyperprolactinemia resulting from disturbance of the pituitary stalk and loss of tonic inhibition of PRL “stalk effect”). Visual changes, including diminished acuity or visual field deficits, may result from tumor compression of the optic apparatus or hemorrhage (Fig. 5). Mass effect producing increased intracranial pressure may evoke headache, nausea, vomiting and papilledema. Memory problems and behavioral changes may also be seen. All patients suspected of harboring a tumor in the region of the sella turcica should undergo a complete neurological, ophthalmological, endocrinological, and radiological workup. A neurological examination is performed noting any focal neurological deficits including cranial neuropathies. All patients old enough to cooperate should undergo formal visual field testing, acuity testing, and dilated fundoscopic examination. Each facet of the hypothalamic-pituitary-end organ axis should be assessed, including diabetes insipidus. Serum PRL levels should be evaluated in all patients with pituitary tumors. Mild elevation may be due to a “stalk effect” (loss of tonic inhibition) while levels greater than 200 µg/mL support the presence of a PRL-secreting adenoma. Thyroid function is evaluated by measuring free thyroxine, thyroxine, and thyroid-stimulating hormones. Adrenal function is assessed by a morning serum cortisol measurement. In the instance of suspected Cushing disease, 24-hour urine-free cortisol is evaluated (age permitting) and a dexamethasone suppression tests can be performed. To evaluate for growth hormone status, serum GH and insulin-like growth factor-1 (IGF-1) levels are measured. In children, a radiograph can be obtained to assess bone age in comparison with chronological age. An oral glucose tolerance test with GH-determinations can be performed when possible in cases of suspected growth-hormone-secreting tumors. Radiological imaging is achieved with dedicated MR imaging of the sellar region. At times, a computed tomographic (CT) scan may be useful to assess the degree of aeration of the sella, particularly in younger patients where the sella has not yet become fully pneumatized. CT can also be performed to help differentiate a Rathke cleft cyst
358
Jagannathan et al.
from craniopharyngioma in patients who present with cystic sellar/parasellar lesions, as well as to identify bony abnormalities (such as sphenoid septations, bone dehiscence) prior to surgery. Neurological disturbances, such as headache and visual field defects, along with manifestations of endocrine deficiency are the common presenting symptoms of sellar region tumors. These tumors can often stretch the diaphragma sellae and cause headaches (41,42). Obstruction of the cerebral aqueduct and the foramen of Monro may also occur, making a shunt necessary. Large tumors with lateral growth may also compress the temporal lobes and cause seizures. At diagnosis, endocrine dysfunction is found in many patients (43,44). Reduced GH secretion is the most frequent endocrinopathy and can be present in up to 75% of patients. This is followed by follicle stimulating hormone/luteinizing hormone (FSH/LH) deficiency, which can be seen in 40% of patients, and then ACTH and TSH deficiency in 25% (21). Both premature sexual development in the first decade of life and pubertal delay in adolescent patients can be seen. Diminished libido and loss of vitality are common symptoms, especially in males.
PRL-Secreting Adenomas In the absence of complications necessitating immediate surgery, such as rapidly progressing visual loss, hydrocephalus, or cerebrospinal fluid leak, pharmacotherapy with dopamine agonists should be considered the first-line treatment approach. Dopamine agonists, including bromocriptine and cabergoline, effectively normalize PRL levels in as many as 89% of patients (58–60). These medications are not only effective biochemically but tumor volume also decreases by at least 50% in more than two-thirds of patients within the first few months of therapy, and, more importantly, visual field deficits improve in all but 10% of patients (61). Transsphenoidal surgery is most successful in obtaining remission in the setting of microprolactinomas. In this population, PRL levels can be normalized in 50% to 90%, with experienced centers reporting results around 85%. Not unexpectedly, results with macroprolactinomas are less successful. Surgical remission may be expected in 28% to 56%, with most experienced centers reporting remission in about half the patients (17,35,62–65).
Cushing Disease MEDICAL AND SURGICAL TREATMENT OF SELLAR REGION TUMORS Pituitary Adenomas Clinically Nonfunctioning Adenomas The first approach to these adenomas is transsphenoidal resection to debulk the tumor and decompress parasellar/suprasellar structures. As in the other adenoma histotypes, surgery has a low morbidity and leads to an improvement of visual symptoms in the majority of cases (17,35). New endocrine deficits, seen more frequently in macroadenomas, have been reported in up to 40% (45). However, recent results indicate that only 3% of patients with microadenomas and 5% of patients with macroadenomas with preoperative normal pituitary function experience new hormonal deficits following surgery (46). Immediate postoperative polyuria occurs in about 30% of patients, but in only 3% to 10% does this polyuria persist beyond the first week of surgery (47). Delayed hyponatremia, occurring most often 7 to 10 days after surgery, is evident in 1% to 2.4%. Worsening in preoperative vision can be seen in 1% to 4% (48–50). Anatomic complications include nasal septal perforations in 7% and fat graft hematomas in 4%. Postoperative cerebrospinal fluid leaks and meningitis are reported in 0.5% to 3.9% (46,51–53). For tumors with incomplete resection, radiosurgery and medical and radiation therapy can be considered. Neither medical therapy nor radiation therapy is usually recommended as primary treatment, as the long-term effects of medical management are unknown and radiation therapy has a long latency before effective control can be seen. The recent development of the endoscopic transsphenoidal approach to the pituitary region (54), which has similar indications to conventional transsphenoidal microsurgery, offers some potential advantages over traditional surgical approaches due to its minimal invasiveness and panoramic visualization. This procedure requires no submucosal dissection and therefore no postoperative nasal packing. Furthermore, a wider panoramic surgical view of the operating field is obtained, which potentially improves the likelihood of a more complete and safer tumor removal (see section below). Endoscopic treatments can result in shorter hospitalization and a rapid recovery in pediatric patients (54–57).
Transsphenoidal resection is the treatment of choice for ACTH-secreting adenomas. Surgical excision is successful in the majority of children, with initial remission rates of 70% to 98% and long-term remission of 50% to 98% in most studies (52,66–73). The success rate decreases when the patients are followed up for more than 5 years (74), and the outcome cannot be accurately predicted by preoperative tests (75,76). The morbidity is low when the procedure is carried out by an experienced team (53). Transsphenoidal microsurgery is considered successful when it is followed by remission of signs and symptoms of hypercortisolism and by normalization of laboratory values. Successful surgery is usually followed by adrenal insufficiency and patients may require hydrocortisone replacement for 6 to 12 months. In pediatric patients, resumption of normal growth or even catch-up growth can be observed. Generally, final height is compromised compared to target height (2). However, some children do achieve a normal final stature. The optimal treatment modality in patients who have relapses after transsphenoidal adenomectomy is still controversial. Some authors recommend repeat surgery (68,69), while others favor radiation therapy or radiosurgery (77). Although surgery can induce panhypopituitarism or permanent diabetes insipidus, hypothalamic-pituitary dysfunction is a relatively frequent ultimate complication of radiation (78). Bilateral adrenalectomy may be the last therapeutic option in case of failure of both surgery and radiotherapy. Stereotactic radiosurgery with the Gamma Knife, Cyberknife, proton beam, or a linear accelerator is a promising modality that minimizes the toxic effects of radiation on the brain, while still controlling tumor growth and ACTH secretion. Caution must be used in the pediatric population however, as the long-term effects of radiation in children are largely unknown.
GH-Secreting Adenomas The objectives of treatment of GH excess are tumor removal with resolution of its mass effect, restoration of normal basal and stimulated GH secretion, relief of symptoms caused by GH excess, and prevention of the disease sequelae (i.e., hypertension, insulin resistance, diabetes mellitus, and lipid abnormalities) (79). The currently available treatment options for acromegaly include surgery, radiotherapy,
Chapter 23: Evaluation and Management of Sellar Tumors
and pharmacological suppression of GH levels by means of DA-agonists somatostatin analogues or GH-receptor antagonists (17,20,26,32,34,74,80,81). Although medical therapy is increasingly improving, transsphenoidal surgery remains the first-line therapy for GH adenomas (51,79,82). Surgery can achieve rapid biochemical remission (normal IGF-1 levels, nadir GH < 1 µg/L during oral glucose tolerance test (OGTT)) in about 85% of patients with microadenomas and 50% of those with macroadenomas (25), and in pediatric patients with gigantism, transsphenoidal surgery was found to be as safe as in adults (23,26). The success rate of surgery is further improved when performed by a surgeon who specializes in transsphenoidal surgery. Once in remission, about 8% of patients recur at 10 years (23,26). Of these patients, 48% can achieve remission with repeated transsphenoidal surgery (81). Craniopharyngiomas Neurological disturbances, such as headache and visual field defects, as well as endocrine deficiencies such as growth retardation and delayed puberty are common presenting symptoms of craniopharyngiomas. These tumors can stretch the diaphragma sellae and can cause headaches and obstructive hydrocephalus (38,39). At diagnosis, endocrine dysfunction is found in up to 80% of patients (83,84). Reduced GH secretion is the most frequent endocrinopathy and can be present in up to 75% of patients. This is followed by FSH/LH deficiency, which can be seen in 40% of patients, and then ACTH and TSH deficiency in 25% (83,84). Despite the fact that craniopharyngiomas are frequently large at presentation, the pituitary stalk is usually not disrupted, and hyperprolactinemia secondary to pituitary stalk compression is found in only 20% of patients. Diabetes insipidus is also relatively uncommon, occurring in less than 20% of patients (83,84). Increased availability of high-resolution MR imaging has greatly improved the visualization and radiological diagnosis of craniopharyngiomas. The neuroradiological diagnosis of craniopharyngiomas is based on the features of the lesion itself and on its relationship with the surrounding structures. The diagnosis is mainly based on the three characteristic components of the tumor: cystic, solid, and calcified. The cystic component constitutes the most important part of the tumor, and shows variable signal depending on the chemical–physical properties of its content. A fluid content will appear hypointense in T1-weighted and hyperintense in T2-weighted images, whereas a lipid, methemoglobin, or protein content will appear as hyperintense in T1 and T2 sequences (85,86). In small intrasellar or enclosed tumors, total resection is most easily achieved, and adjunctive radiotherapy is unnecessary (1–2). Radiotherapy is required in cases of incomplete tumor removal, which occurs frequently with extrasellar craniopharyngiomas (the majority of cases) (87). Surgical morbidity of craniopharyngiomas depends on tumor size and invasiveness, the experience of the surgeon, and the route of surgical approach. The risk of hypothalamic damage is significantly greater in large invasive tumors treated via the transcranial approach (88–90). Near-total excision of the tumor by an experienced surgeon sparing the hypothalamus, carotid arteries, and visual apparatus, followed by fractionated radiotherapy, provides the best hope of low long-term morbidity and longer survival. Regardless of the approach, the incidence of endocrine dysfunction is high following surgical treatment, although it is lower after the transsphenoidal approach (91). Localized intracavity yttrium, phosphorus-32,
359
and other radioactive implants, given as additional treatment, have proven useful for recurrent tumors with a predominant cystic component (80,92,93).
TRANSSPHENOIDAL SURGERY Microscopic transsphenoidal surgery is the treatment of choice for patients with sellar tumors, with the exception of prolactinomas, and is the first treatment for the majority of patients (74,94). Over time, we have gradually evolved to use the endoscopic approach in patients with pituitary adenomas, although we have had great success with the microscope as well (95–98). Our standard approach now is to use a binasal three- or four-hand technique (in lieu of an endoscope holder), with a posterior septectomy preserving the middle turbinates unless the nasal cavity is exceptionally small (54). The advantage of the endoscopic exposure is a broad exposure of the sella, permitting visualization and examination of the pituitary gland. Patients are positioned in a semirecumbent position (approximately 20 degrees back up) with the head placed in a horseshoe headrest (Fig. 6). Patties soaked in oxymetazoline are placed between the middle turbinates and the septum prior to preparation and draping. After draping, the patties are removed, and using a 25-gauge 1.5-inch needle, or spinal needle, the middle turbinate, posterior portions of the septum and rostrum, as well as the region around the sphenopalatine foramen are injected submucosally with 0.2% ropivacaine with 1:200,000 epinephrine. The 0-degree endoscope is first used to inspect both nasal cavities. The side with the greater working room is chosen for the initial dissection of the posterior septum and sphenoidotomy (if there is a deviated septum, for example, the wider nasal cavity is used). If the sides are equal, our bias is to perform the majority of the dissection in the right naris. Prior to beginning the dissection, the endonasal anatomy is identified, including the middle and inferior turbinates, the choana, and, most important, the superior turbinates and sphenoid ostia. This initial orientation helps to prevent the undesirable breach of the anterior cranial fossa, as there is a tendency to deviate superiorly when inexperienced with the endoscope.
Figure 6 Patients are typically placed in a semirecumbent position prior to surgery with elevation of the head of bed to approximately 20 degrees to avoid venous congestion. Mayfield pin fixation is used in cases (such as this) where neuronavigation is used. (Reproduced with permission) (99).
360
Jagannathan et al.
Figure 7 The panoramic view offered by the endoscope allows inspection of the tumor bed postoperatively to ensure the adequacy of the resection.
The middle turbinates are lateralized and the sphenoid ostium is identified medial to the inferior third portion of the superior turbinate. The mucosa surrounding the sphenoid ostium is incised and the septum is reflected into the contralateral nasal cavity by detaching it from its junction with the sphenoid rostrum. This allows the contralateral sphenoid ostium to be identified and the bone between the two ostia is removed. This initial sphenoidotomy is then enlarged and a posterior septectomy is performed. We have found that the soft-tissue shaver provides a rapid and precise removal of the septal mucosa. A wide anterior sphenoidotomy is performed. The superior limit should allow visualization of the planum sphenoidale, the optic protuberances, and the opticocarotid recesses. During tumor resection, the endoscope will be positioned superolaterally; therefore, the sphenoidotomy will need to be large enough to accommodate the endoscope without obstructing the flow of instruments. The inferior limits of the sphenoidotomy must allow the suction to pass freely to the floor of the sella. In general, the vomer does not need to be completely removed to provide this exposure, and its presence helps establish an anatomical midline if reoperation is necessary. The panoramic view of the sphenoid anatomy provides the information regarding the midline and limits of the sella (Fig. 7). Under endoscopic guidance, the vomer and the midpoint between the carotid protuberances or the opticocarotid recesses can be used as midline markers. Once the intersphenoid sinus septae are removed, the carotid and optic protuberances are identified, as are the opticocarotid recesses, clivus, sella, and planum. We have found that all of these landmarks are not always visible. Nevertheless, enough of the anatomical landmarks are usually visible so that the sellar anatomy can be determined. If doubt continues, intraoperative video fluoroscopy is performed. In general, the posterior septectomy and sphenoidotomy are performed using a two-handed mono-nostril technique, with the endoscope in one hand and an operative instrument in the other. The opposite nostril is accessed at times, particularly during the septectomy, but the majority of the endoscopy and dissection is performed via one nostril. From this point forward, the operation proceeds through a binasal three- or four-hand approach, with the endoscope and suction in one nostril and the operative instrument in
Figure 8 (Left panel) Endoscopic perspective of the entire sella before opening the floor. The endoscope allows the surgeon to visualize the opticocarotid recesses (white arrow), thus facilitating a safe opening. (Right panel) Using the endoscope, the interface between the gland and the adenoma can be well visualized as demonstrated here. The adenoma is then gently dissected away from the gland.
the other. Interplay is required between the endoscopist and the surgeon. The endoscope should follow the instruments into and out of the nose and dynamically focus on the field of interest. The sellar floor is then opened to the limits of the cavernous sinus laterally, the tuberculum and planum superiorly, and the sellar floor inferiorly (Fig. 8). We have not found it necessary to use a drill for the sellar opening, and we often simply fracture the sellar floor or open it with a fine chisel. At times, a small vessel can be seen to exit the dura in the midline. We have found that this vessel correlates well with the anatomical midline and can be used in addition to the other midline markers. The 0-degree endoscope is used during the majority of the operation. ACTH adenomas require a different resection from other tumors. The key to tumor resection involves obtaining a bloodless surgical field. When the dura is opened, the surface of the pituitary gland is examined thoroughly for the presence of tumor. If tumor is identified, an attempt is made to microdissect the entire capsule of the tumor. If an extracapsular dissection is not possible, the tumor resection proceeds in a manner similar to that of the microscopic adenomectomy, with a removal of the inferior, then the lateral, and finally the superior portions of the tumor against the diaphragm. If no tumor is immediately visualized, a horizontal incision is made at the equator of the exposed gland. Then, a vertical paramedian incision is made with the dissection proceeding in a progressive posterior and lateral direction in an attempt to identify a tumor. In select patients, with negative MR imaging, if no tumor is identified, then the same maneuver is performed via a paramedian incision on the contralateral side. If still no tumor is identified, then the inferior mesial portion of the gland (the “mucoid wedge”) is dissected via the transverse incision leaving a superior central rim of gland still attached to the pituitary stalk. A hemihypophysectomy may be done when petrosal sinus sampling is convincingly lateralized. The operation is typically considered complete at this stage in pediatric patients and women of childbearing age who desire fertility. In other patients, the decision to do a hypophysectomy is made on a case-by-case basis depending on the severity of the patient’s condition, results of prior procedures, and results of laboratory studies and IPSS.
Chapter 23: Evaluation and Management of Sellar Tumors
The differences from the traditional microscopic approach are the more magnified views of areas that are out of a microscope’s line of sight and the ability to move the endoscope close in for dynamic views of areas of interest. With the 0- and 30-degree endoscopes, the sellar floor, cavernous sinus walls, and diaphragm can be directly inspected. When the anatomy is not completely visualized, the 30-degree endoscope can be used. In some situations intraoperative ultrasound may also be of utility in clarifying the relationship to the cavernous sinus and occasionally in determining the presence of tumor (100). Patients with Cushing disease who do not achieve remission at surgery deserve mention. Repeat surgery for a total hypophysectomy is considered for some adult patients beyond their childbearing years in whom a tumor was identified pathologically but who did not achieve remission. Radiosurgery is considered for patients in whom no tumor was identified at surgery. Oldfield et al. (101) have reported on the use of multiple blind incisions in the pituitary gland in these cases to infarct small tumors that may be present, a strategy that may be useful in some patients. If a CSF leak is noted during surgery, we place a fat graft within the sella. Appropriately sized fat grafts can be difficult to place within the nose, and at times a nasal speculum must be used to provide access. Often we will also place a dural substitute against the diaphragm and in a subdural plane over the fat. If no CSF leak is noted, we simply place Gelfoam within the sella. If adequate septal bone is not harvested during the procedure (as is the norm for endoscopic approaches), a tailored artificial graft is placed in an extradural plane under the remaining sellar bone. In the case of a CSF leak, we usually do not fill the sphenoid sinus with fat; if we do, however, we first remove the sphenoid mucosa. Otherwise, every attempt is made to preserve the sphenoid mucosa, and it is disrupted over the sella only.
Extended Transsphenoidal Approaches A modification of the traditional transsphenoidal approach was described by Weiss et al. in 1987 (102). This approach has been termed “the extended transsphenoidal approach” and entails removal of additional bone along the tuberculum sellae and the posterior planum sphenoidale with subsequent opening of the dura mater above the diaphragma sellae (Fig. 9). This route allows excellent midline access and visibility to the suprasellar space while obviating brain retraction. The extended transsphenoidal approach allows for additional exposure of the suprasellar space, and is useful for a variety of sellar region pathologies, including tuberculum sellae meningiomas, craniopharyngiomas, and supraglandular Rathke cysts, which previously were approached only via a craniotomy. The major advantages of the extended transsphenoidal approach are that the need for brain retraction is obviated; manipulation of the optic apparatus is minimized; and early identification of the pituitary gland and infundibulum is allowed, increasing the likelihood of preserving pituitary function. The challenges to the extended transsphenoidal approach are the selection of suitable patients. Not every patient with a midline suprasellar lesion is a candidate, and the approach depends on the surgical anatomy, which can be determined by preoperative MR imaging and can serve as a road map to determining if an extended transsphenoidal approach will be safe and adequate for dealing with a given lesion. The
361
goals of surgery must be evaluated carefully as well, and for some cases in which palliation is the goal, the technique is quite safe and effective. In cases in which gross-total removal is the goal, the difficulties of dealing with the blood supply of the tumor and its adherence (if any) to the optic apparatus and/or hypothalamus should also be considered. Our published experience with the extended transsphenoidal skull base technique at the University of Virginia included 56 patients who underwent surgery between 1999 and 2005 (103,104). Forty-eight of these patients had lesions other than pituitary adenomas. They included 30 craniopharyngiomas, three entirely suprasellar Rathke cleft cysts, seven meningiomas, two germ cell tumors, one granular cell tumor, a salivary cyst, a chordoma of the superior clivus, a granulomatous lesion of the pituitary stalk, a case of solid lymphocytic hypophysitis involving the optic chiasm, a hemangioblastoma, and a chordoid glioma. Surgery was generally accomplished in a satisfactory fashion. There were two meningiomas that could not be removed: one because it was entirely suprasellar and without a dural attachment, and the other because it was simply too large and too firm to be mobilized and removed safely. One craniopharyngioma was not removed because of profuse bleeding from a very elaborate intercavernous sinus. One patient with a craniopharyngioma died postoperatively of a hemorrhage from the apex of the tumor. We were able to accomplish gross-total removal of 25 craniopharyngiomas, all three Rathke cysts, and five meningiomas. Twenty-three patients (64%) in this series had visual improvement after surgery, 5 patients (14%) had transient visual decline, and 3 patients (8%) had permanent worsening of vision. The extended transsphenoidal approach requires a large opening in the dura mater over the tuberculum sellae and the posterior planum sphenoidale and typically results in large intraoperative CSF leaks, which necessitate precise and effective dural closure to prevent a postoperative CSF fistula and meningitis.
Figure 9 Saggital view of the extended transsphenoidal approach. By removing additional bone along the tuberculum sellae and the posterior planum sphenoidale, excellent midline access and visibility of the suprasellar space is provided, while minimizing brain retraction.
362
Jagannathan et al.
SELLAR AND NASAL RECONSTRUCTION Following tumor removal and hemostasis, the sella must be carefully inspected to allow reconstruction and closure. We prefer not to leave dead space in the sella, particularly in cases with a CSF leak. If no CSF leak has occurred, the sella can be loosely packed with Gelfoam, but in the event of a CSF leak, a tissue graft is often necessary. Sellar floor reconstruction can be performed with a piece of homologous dural graft material or fascia lata, but this is unlikely to augment the closure. Our current practice is to pack the sella with fat taken from a periumbilical incision. The fat is soaked in 10% chloramphenicol solution and may be rolled in Avitene before being trimmed to the appropriate size and placed in the sella. The sellar floor is then reconstructed with a trimmed piece of cartilage or bone or an artificial plate (MedPor). In cases of repeat transsphenoidal surgery, alternative materials may be used such as an iliac crest bone graft or methylmethacrylate. Some surgeons have also reported on the use of fibrin glue or other sealants as a part of the closure in patients with CSF leaks, although we have not found this to be reliably helpful in our practice (91,105). During the nasal closure, careful attention must be paid to restoring the anatomic and physiologic features of the nose. The septal flaps should be reapproximated and the nasal septum should be returned to its midline position. Mucosal tears may also be repaired using fine catgut sutures, particularly when a sublabial or hemitransfixion approach is used.
IMMEDIATE POSTOPERATIVE CARE Most patients are extubated, awake, and alert following transsphenoidal surgery. We do not routinely use a urinary catheter or intensive care unit monitoring following routine transsphenoidal cases. In all patients, however, vigilant postoperative monitoring of visual function, fluid intake and output, and electrolyte balance is essential. Some degree of diuresis does occur during the postoperative period, and does not necessarily imply a diagnosis of diabetes insipidus; however, making the diagnosis of true diabetes insipidus, which is characterized by a brisk diuresis, along with alterations in the serum/urine electrolyte levels is essential. Fortunately, diabetes insipidus when it does occur is normally transient. In a patient who is awake and able to keep up with fluid losses by drinking, treatment with desmopressin (DDAVP) may not be necessary. However, in patients with neurological deficit, DDAVP supplementation along with adequate fluid intake is critical. Following a stress dose of steroids, exogenous steroids are normally rapidly tapered after the first postoperative day. If there is any doubt about the function of the hypothalamicpituitary-adrenal axis, steroid replacement in physiologic doses should be administered until formal endocrine testing is allowed.
COMPLICATIONS AND AVOIDANCE The transsphenoidal approach is usually a safe procedure with a low complication rate. Since the procedure lacks visible scars and has a low morbidity compared with open surgery, our general experience is that patients’ postoperative satisfaction is high. Vascular injury is a rare but wellknown complication following transsphenoidal surgery, and is one of the main sources of operative mortality. The intracavernous portion of the internal carotid artery appears to be the most sensitive to injury, followed by the circle of Willis.
The sequelae of these injuries are usually intracranial hemorrhage, thrombotic stroke, pseudoaneurysm formation, and carotid-cavernous fistulae formation. When vascular injury is suspected intraoperatively, tamponade should be used to control hemorrhage, and a postoperative angiogram should be obtained. CSF rhinorrhea and meningitis are among the more common, serious complications associated with transsphenoidal surgery. These are the result of disruption of the diaphragma sellae, which can be thin and is often adherent to tumor. In most instances, good microsurgical technique can avoid these complications, but in other cases CSF leakage is inevitable. In cases of CSF leak, a careful sellar closure is critical (method outlined above). Early recognition of a postoperative leak is also important in reducing the risk for infection. Our preference is not to perform lumbar drainage in patients with postoperative CSF leaks; we prefer to perform an immediate re-exploration of the sella with repacking and reconstruction of the floor. Nasal complications are common postoperative complaints following transsphenoidal surgery, particularly after microscopic procedures, where more trauma to the nasal mucosa is incurred. Although rarely fatal, these findings can be irritating to patients and persist for months after surgery. One of the most common causes of nasal irritation is the incorrect use of the retractor, with resulting diastasis or fracture of the hard palate and cribriform plate. In the latter case, CSF leakage may also result. Sphenoid mucosa may also become infected resulting in a febrile sinusitis and the development of a mucocele. Technical errors in handling the nasal mucosa and nasal septum may also result in septal perforations or external nasal deformities, which may be unsightly. Loss of smell has also been noted to occur due to damage of nerve endings in the nasal mucosa.
FOLLOW-UP Patients being treated for pituitary adenomas must be followed for long term generally with serial clinical, ophthalmological, endocrinological, and imaging evaluations. In children, particularly, height, weight, and pubertal status must be carefully monitored in relevant age groups. Tanner staging, skeletal maturation, LH, FSH, and sex hormone levels should also be asessed serially by an endocrinologist. All patients with visual loss should have serial visual field examinations and screening. Patient hormonal status (screening for hypothyroidism, adrenal insufficiency, diabetes insipidus, etc.) must be assessed. Serial testing of thyroid function and GH status (with IGF-1 levels and provocative testing when applicable) should be undertaken. Patients with functioning tumors should be investigated as appropriate (e.g., 24-hour urine-free cortisol testing in patients with Cushing disease). Patients already on hormonal replacement should have their replacement therapy adjusted as necessary. Finally, serial MR imaging should be performed to assess for tumor recurrence. Generally, an initial postoperative study is performed 6 weeks to 3 months following treatment and repeated yearly thereafter (or more frequently as indicated). REFERENCES 1. Jagannathan J, Dumont AS, Jane JA Jr. Diagnosis and management of pediatric sellar lesions. Front Horm Res. 2006;34:83–104. 2. Jagannathan J, Dumont AS, Jane JA Jr, et al. Pediatric sellar tumors: Diagnostic procedures and management. Neurosurg Focus. 2005;18:E6.
Chapter 23: Evaluation and Management of Sellar Tumors 3. Post KD, McCormick PC, Hays AP, et al. Metastatic carcinoma to pituitary adenoma. Report of two cases. Surg Neurol. 1988;30:286–292. 4. Post KD. Supplementary comment on “Endonasal endoscopic transsphenoidal chiasmapexy with silicone plates for empty sella syndrome–technical note–.” Neurol Med Chir (Tokyo). 2005;45:491. 5. Castillo M. Pituitary gland: Development, normal appearances, and magnetic resonance imaging protocols. Top Magn Reson Imaging. 2005;16:259–268. 6. Escourolle H, Abecassis JP, Bertagna X, et al. Comparison of computerized tomography and magnetic resonance imaging for the examination of the pituitary gland in patients with Cushing’s disease. Clin Endocrinol (Oxf). 1993;39:307–313. 7. Findling JW, Doppman JL. Biochemical and radiologic diagnosis of Cushing’s syndrome. Endocrinol Metab Clin North Am. 1994;23:511–537. 8. Sheehan JM, Lopes MB, Sheehan JP, et al. Results of transsphenoidal surgery for Cushing’s disease in patients with no histologically confirmed tumor. Neurosurgery. 2000;47:33–36; discussion 37–39. 9. Buchfelder M, Nistor R, Fahlbusch R, et al. The accuracy of CT and MR evaluation of the sella turcica for detection of adrenocorticotropic hormone-secreting adenomas in Cushing disease. AJNR Am J Neuroradiol. 1993;14:1183–1190. 10. Arnaldi G, Angeli A, Atkinson AB, et al. Diagnosis and complications of Cushing’s syndrome: A consensus statement. J Clin Endocrinol Metab. 2003;88:5593–5602. 11. Patronas N, Bulakbasi N, Stratakis CA, et al. Spoiled gradient recalled acquisition in the steady state technique is superior to conventional postcontrast spin echo technique for magnetic resonance imaging detection of adrenocorticotropin-secreting pituitary tumors. J Clin Endocrinol Metab. 2003;88:1565–1569. 12. Batista D, Courkoutsakis NA, Oldfield EH, et al. Detection of adrenocorticotropin-secreting pituitary adenomas by magnetic resonance imaging in children and adolescents with cushing disease. J Clin Endocrinol Metab. 2005;90:5134–5140. 13. Jagannathan J, Sheehan JP, Jane JA Jr. Evaluation and management of Cushing syndrome in cases of negative sellar magnetic resonance imaging. Neurosurg Focus. 2007;23:E3. 14. Davis WL, Lee JN, King BD, et al. Dynamic contrast-enhanced MR imaging of the pituitary gland with fast spin-echo technique. J Magn Reson Imaging. 1994;4:509–511. 15. Sakamoto Y, Takahashi M, Korogi Y, et al. Normal and abnormal pituitary glands: Gadopentetate dimeglumine-enhanced MR imaging. Radiology. 1991;178:441–445. 16. Stadnik T, Stevenaert A, Beckers A, et al. Contrast behavior between microadenoma and normal pituitary gland after gadolinium injection as a function of time at 1.5 T. Neuroradiology. 1992;34:184–189. 17. Laws ER, Sheehan JP. PituitarySsurgery: A Modern Approach. Basel, Switzerland, New York: Karger, 2006. 18. Ezzat S, Asa SL, Couldwell WT, et al. The prevalence of pituitary adenomas: A systematic review. Cancer. 2004;101:613–619. 19. Johnson MD, Woodburn CJ, Vance ML. Quality of life in patients with a pituitary adenoma. Pituitary. 2003;6:81–87. 20. Lindholm J. Cushing’s syndrome: Historical aspects. Pituitary. 2000;3:97–104. 21. Kov´acs K, Horvath E, Universities Associated for Research and Education in Pathology. Tumors of the pituitary gland. Washington, D.C., Armed Forces Institute of Pathology, 1986. 22. Hamilton CR Jr, Maloof F. Acromegaly and toxic goiter. Cure of the hyperthyroidism and acromegaly by proton-beam partial hypophysectomy. J Clin Endocrinol Metab. 1972;35:659–664. 23. Laws ER Jr. Acromegaly and gigantism. In: Wilkins RH RS, ed. Neurosurgery. New York: McGraw Hill, 1985:864–867. 24. Katznelson L, Kleinberg D, Vance ML, et al. Hypogonadism in patients with acromegaly: Data from the multicentre acromegaly registry pilot study. Clin Endocrinol (Oxf). 2001;54:183–188. 25. Kreutzer J, Vance ML, Lopes MB, et al. Surgical management of GH-secreting pituitary adenomas: An outcome study
26. 27. 28. 29.
30. 31.
32.
33. 34.
35.
36. 37.
38.
39.
40. 41.
42.
43.
44.
45.
46.
47.
48.
49. 50.
363
using modern remission criteria. J Clin Endocrinol Metab. 2001;86:4072–4077. Laws ER, Vance ML, Thapar K. Pituitary surgery for the management of acromegaly. Horm Res. 2000;53(Suppl 3):71–75. Vance ML. Endocrinological evaluation of acromegaly. J Neurosurg. 1998;89:499–500. Ezzat S. Hepatobiliary and gastrointestinal manifestations of acromegaly. Dig Dis. 1992;10:173–180. Ladas SD, Thalassinos NC, Ioannides G, et al. Does acromegaly really predispose to an increased prevalence of gastrointestinal tumours? Clin Endocrinol (Oxf). 1994;41:597–601. Molitch ME. Clinical manifestations of acromegaly. Endocrinol Metab Clin North Am. 1992;21:597–614. Vasen HF, van Erpecum KJ, Roelfsema F, et al. Increased prevalence of colonic adenomas in patients with acromegaly. Eur J Endocrinol. 1994;131:235–237. Melmed S, Jackson I, Kleinberg D, et al. Current treatment guidelines for acromegaly. J Clin Endocrinol Metab. 1998;83:2646–2652. Foo TK, Khooi TK. Re: Acromegaly and cerebrovascular accidents. Singapore Med J. 1989;30:226. Louwerens M, de Herder WW, Postema PT, et al. Pituitary insufficiency and regression of acromegaly caused by pituitary apoplexy following cerebral angiography. Eur J Endocrinol. 1996;134:737–740. Laws ER Jr, K T. Recurrent pituitary adenomas. In: Vance ML, ed. Pituitary Adenomas. Edinburgh, UK: Churchill-Livingtone, 1996:385–394. Vance ML. Medical treatment of functional pituitary tumors. Neurosurg Clin N Am. 2003;14:81–87. Bunin GR, Surawicz TS, Witman PA, et al. The descriptive epidemiology of craniopharyngioma. Neurosurg Focus. 1997; 3:e1. Haupt R, Magnani C, Pavanello M, et al. Epidemiological aspects of craniopharyngioma. J Pediatr Endocrinol Metab. 2006;19(Suppl 1):289–293. Bunin GR, Surawicz TS, Witman PA, et al. The descriptive epidemiology of craniopharyngioma. J Neurosurg. 1998;89:547– 551. Akimura T, Kameda H, Abiko S, et al. Infrasellar craniopharyngioma. Neuroradiology. 1989;31:180–183. Ebersold MJ, Quast LM, Laws ER Jr, et al. Long-term results in transsphenoidal removal of nonfunctioning pituitary adenomas. J Neurosurg. 1986;64:713–719. Hooper MJ, Stiel JN, Selby G, et al. Clinical features of nonfunctioning chromophobe adenomas. Aust N Z J Med. 1973;3:371– 375. Cannavo S, Venturino M, Curto L, et al. Clinical presentation and outcome of pituitary adenomas in teenagers. Clin Endocrinol (Oxf). 2003;58:519–527. Harms E, Siggelkow H, Buchfelder M, et al. Macroadenoma of the pituitary gland with moderate hyperprolactinaemia. Dtsch Med Wochenschr. 2003;128:667–670. Mortini P, Barzaghi R, Losa M, et al. Surgical treatment of giant pituitary adenomas: Strategies and results in a series of 95 consecutive patients. Neurosurgery. 2007;60:993–1002; discussion 1003–1004. Barker FG II, Klibanski A, Swearingen B. Transsphenoidal surgery for pituitary tumors in the United States, 1996-2000: Mortality, morbidity, and the effects of hospital and surgeon volume. J Clin Endocrinol Metab. 2003;88:4709–4719. Nicola G, Tonnarelli G, Griner A. Surgery for recurrence of pituitary adenomas. In: G F, Beck-Peccoz P, eds. Pituitary Adenomas: New Trends in Basic and Clinical Research. Amsterdam: Excerpta Medica, 1991:329–338. Adams CB. The management of pituitary tumours and post-operative visual deterioration. Acta Neurochir (Wien). 1988;94:103–116. Kaur A, Banerji D, Kumar D, et al. Visual status in suprasellar pituitary tumours. Indian J Ophthalmol. 1995;43:131–134. Mohr G, Hardy J, Comtois R, et al. Surgical management of giant pituitary adenomas. Can J Neurol Sci. 1990;17:62–66.
364
Jagannathan et al.
51. Buchfelder M, Fahlbusch R, Schott W, et al. Long-term followup results in hormonally active pituitary adenomas after primary successful transsphenoidal surgery. Acta Neurochir Suppl (Wien). 1991;53:72–76. 52. Chang CZ, Wang CJ, Howng SL. Pituitary adenomas in adolescence—ten-year experience and literature review. Kaohsiung J Med Sci. 1999;15:691–696. 53. Ciric I, Ragin A, Baumgartner C, et al. Complications of transsphenoidal surgery: Results of a national survey, review of the literature, and personal experience. Neurosurgery. 1997;40:225–236; discussion 236–227. 54. Jane JA Jr, Han J, Prevedello DM, et al. Perspectives on endoscopic transsphenoidal surgery. Neurosurg Focus. 2005; 19:E2. 55. Cappabianca P, Cavallo LM, Esposito F, et al. Endoscopic endonasal transsphenoidal surgery: Procedure, endoscopic equipment and instrumentation. Childs Nerv Syst. 2004;20:796–801. 56. Dedivitis RA, Guimaraes AV. Preliminary results from minimally invasive video-assisted thyroidectomy. Sao Paulo Med J. 2005;123:298–299. 57. Dedivitis RA, Guimaraes AV, Guirado CR. Outcome after partial frontolateral laryngectomy. Int Surg. 2005;90:113–118. 58. Colao A, Di Sarno A, Landi ML, et al. Macroprolactinoma shrinkage during cabergoline treatment is greater in naive patients than in patients pretreated with other dopamine agonists: A prospective study in 110 patients. J Clin Endocrinol Metab. 2000;85:2247–2252. 59. Colao A, di Sarno A, Pivonello R, et al. Dopamine receptor agonists for treating prolactinomas. Expert Opin Investig Drugs. 2002;11:787–800. 60. Corsello SM, Ubertini G, Altomare M, et al. Giant prolactinomas in men: Efficacy of cabergoline treatment. Clin Endocrinol (Oxf). 2003;58:662–670. 61. Webster J. Clinical management of prolactinomas. Baillieres Best Pract Res Clin Endocrinol Metab. 1999;13:395–408. 62. Blackwell RE, Bradley EL Jr, Kline LB, et al. Comparison of dopamine agonists in the treatment of hyperprolactinemic syndromes: A multicenter study. Fertil Steril. 1983;39:744–748. 63. Charpentier G, de Plunkett T, Jedynak P, et al. Surgical treatment of prolactinomas. Short- and long-term results, prognostic factors. Horm Res. 1985;22:222–227. 64. Ikeda H, Jokura H, Yoshimoto T. Transsphenoidal surgery and adjuvant gamma knife treatment for growth hormone-secreting pituitary adenoma. J Neurosurg. 2001;95:285–291. 65. Laws ER Jr, Fode NC, Redmond MJ. Transsphenoidal surgery following unsuccessful prior therapy. An assessment of benefits and risks in 158 patients. J Neurosurg. 1985;63:823–829. 66. Chandler WF, Schteingart DE, Lloyd RV, et al. Surgical treatment of Cushing’s disease. J Neurosurg. 1987;66:204–212. 67. Devin JK, Allen GS, Cmelak AJ, et al. The efficacy of linear accelerator radiosurgery in the management of patients with Cushing’s disease. Stereotact Funct Neurosurg. 2004;82:254–262. 68. Devoe DJ, Miller WL, Conte FA, et al. Long-term outcome in children and adolescents after transsphenoidal surgery for Cushing’s disease. J Clin Endocrinol Metab. 1997;82:3196–3202. 69. Dickerman RD, Oldfield EH. Basis of persistent and recurrent Cushing disease: An analysis of findings at repeated pituitary surgery. J Neurosurg. 2002;97:1343–1349. 70. Esposito F, Dusick JR, Cohan P, et al. Clinical review: Early morning cortisol levels as a predictor of remission after transsphenoidal surgery for Cushing’s disease. J Clin Endocrinol Metab. 2006;91:7–13. 71. Estrada J, Boronat M, Mielgo M, et al. The long-term outcome of pituitary irradiation after unsuccessful transsphenoidal surgery in Cushing’s disease. N Engl J Med. 1997;336:172–177. 72. Estrada J, Garcia-Uria J, Lamas C, et al. The complete normalization of the adrenocortical function as the criterion of cure after transsphenoidal surgery for Cushing’s disease. J Clin Endocrinol Metab. 2001;86:5695–5699. 73. Friedman RB, Oldfield EH, Nieman LK, et al. Repeat transsphenoidal surgery for Cushing’s disease. J Neurosurg. 1989;71:520– 527.
74. Mampalam TJ, Tyrrell JB, Wilson CB. Transsphenoidal microsurgery for Cushing disease. A report of 216 cases. Ann Intern Med. 1988;109:487–493. 75. Pouratian N, Prevedello DM, Jagannathan J, et al. Outcomes and management of patients with Cushing’s disease without pathological confirmation of tumor resection after transsphenoidal surgery. J Clin Endocrinol Metab. 2007;92:3383–3388. 76. Ram Z, Nieman LK, Cutler GB Jr, et al. Early repeat surgery for persistent Cushing’s disease. J Neurosurg. 1994;80:37–45. 77. Jagannathan J, Sheehan JP, Pouratian N, et al. Gamma knife surgery for Cushing’s disease. J Neurosurg. 2007;106:980–987. 78. Kim MS, Lee SI, Sim JH. Gamma Knife radiosurgery for functioning pituitary microadenoma. Stereotact Funct Neurosurg. 1999;72(Suppl 1):119–124. 79. Freda PU. How effective are current therapies for acromegaly? Growth Horm IGF Res. 2003;13(Suppl A):S144–S151. 80. Maini CL, Sciuto R, Tofani A, et al. Somatostatin receptor imaging in CNS tumours using 111In-octreotide. Nucl Med Commun. 1995;16:756–766. 81. Melmed S, Vance ML, Barkan AL, et al. Current status and future opportunities for controlling acromegaly. Pituitary. 2002;5:185– 196. 82. Cozzi R, Barausse M, Asnaghi D, et al. Failure of radiotherapy in acromegaly. Eur J Endocrinol. 2001;145:717–726. 83. Sklar CA. Craniopharyngioma: Endocrine abnormalities at presentation. Pediatr Neurosurg. 1994;21(Suppl 1):18–20. 84. Sklar CA. Craniopharyngioma: Endocrine sequelae of treatment. Pediatr Neurosurg. 1994;21(Suppl 1):120–123. 85. Kishore PR, Rao CV, Williams JP, et al. The limitation of computerized tomographic diagnosis of intracranial midline cysts. Surg Neurol. 1980;14:417–431. 86. Richmond IL, Wilson CB. Parasellar tumors in children. I. Clinical presentation, preoperative assessment, and differential diagnosis. Childs Brain. 1980;7:73–84. 87. Laws ER Jr, Vance ML. Radiosurgery for pituitary tumors and craniopharyngiomas. Neurosurg Clin N Am. 1999;10:327– 336. 88. Araki K, Koga M, Okada T, et al. A boy with normal growth in spite of growth hormone deficiency after resection of a suprasellar teratoma. Endocr J. 2000;47(Suppl):S101–S104. 89. Kouri JG, Chen MY, Watson JC, et al. Resection of suprasellar tumors by using a modified transsphenoidal approach. Report of four cases. J Neurosurg. 2000;92:1028–1035. 90. Moreland DB, Diaz-Ordaz E, Czajka GA, et al. Endoscopic resection of pituitary lesions through the nostril. Semin Perioper Nurs. 1998;7:193–199. 91. D’Arrigo C, Landolt AM. Fibrin sealing of mucoperichondrial flaps in endonasal-transsphenoidal pituitary surgery: Technical note. Neurosurgery. 1994;35:529–531; discussion 531–522. 92. Hald JK, Eldevik OP, Skalpe IO. Craniopharyngioma identification by CT and MR imaging at 1.5 T. Acta Radiol. 1995;36:142– 147. 93. Lange M, Kirsch CM, Steude U, et al. Intracavitary treatment of intrasellar cystic craniopharyngeomas with 90-Yttrium by trans-sphenoidal approach—a technical note. Acta Neurochir (Wien). 1995;135:206–209. 94. Invitti C, Pecori Giraldi F, de Martin M, et al. Diagnosis and management of Cushing’s syndrome: Results of an Italian multicentre study. Study Group of the Italian Society of Endocrinology on the Pathophysiology of the Hypothalamic-Pituitary-Adrenal Axis. J Clin Endocrinol Metab. 1999;84:440–448. 95. Carrau RL, Kassam AB, Snyderman CH. Pituitary surgery. Otolaryngol Clin North Am. 2001;34:1143–1155, ix. 96. Huggard D, Khakoo Z, Kassam G, et al. Effect of testosterone on growth hormone gene expression in the goldfish pituitary. Can J Physiol Pharmacol. 1996;74:1039–1046. 97. Kassam A, Snyderman CH, Mintz A, et al. Expanded endonasal approach: The rostrocaudal axis. Part I. Crista galli to the sella turcica. Neurosurg Focus. 2005;19:E3. 98. Kassam A, Snyderman CH, Mintz A, et al. Expanded endonasal approach: The rostrocaudal axis. Part II. Posterior clinoids to the foramen magnum. Neurosurg Focus. 2005;19:E4.
Chapter 23: Evaluation and Management of Sellar Tumors 99. Jagannathan J, Prevedello DM, Ayer VS, et al. Computerassisted frameless stereotaxy in transsphenoidal surgery at a single institution: Review of 176 cases. Neurosurg Focus. 2006;20:E9. 100. Watson JC, Shawker TH, Nieman LK, et al. Localization of pituitary adenomas by using intraoperative ultrasound in patients with Cushing’s disease and no demonstrable pituitary tumor on magnetic resonance imaging. J Neurosurg. 1998;89:927–932. 101. Oldfield EH, Vortmeyer AO. Development of a histological pseudocapsule and its use as a surgical capsule in the excision of pituitary tumors. J Neurosurg. 2006;104:7–19.
365
102. Weiss MH. Transnasal transsphenoidal approach. In: Apuzzo ML, ed. Surgery of the Third Ventricle. Baltimore, MD: Williams & Wilkins, 1987:476–494. 103. Dumont AS, Kanter AS, Jane JA Jr. Extended transsphenoidal approach. Front Horm Res. 2006;34:29–45. 104. Laws ER, Kanter AS, Jane JA Jr, et al. Extended transsphenoidal approach. J Neurosurg. 2005;102:825–827; discussion 827– 828. 105. Van Velthoven V, Clarici G, Auer LM. Fibrin tissue adhesive sealant for the prevention of CSF leakage following transsphenoidal microsurgery. Acta Neurochir (Wien). 1991;109:26–29.
24 Tumors of the Middle Cranial Fossa Ali A. Baaj, Siviero Agazzi, and Harry R. van Loveren
degrees; the internal auditory canal (IAC) bisects this angle. The cochlea lies within the acute angle formed by the IAC and GSPN.
INTRODUCTION Familiarity with the anatomic landmarks of the middle cranial fossa is the first step to understanding the pathology, clinical presentations, and surgical approaches for lesions in this region. The most common tumors of the middle fossa are chordoma, chondrosarcoma, meningioma, trigeminal schwannoma, and osteosarcoma. Diagnosis, imaging studies, preoperative planning, and surgical technique are emphasized. The surgical technique section includes frequently asked questions (FAQs) encountered during numerous meetings and courses where this approach was taught. These FAQs highlight important anatomic landmarks, technical pearls, strategic concepts, and surgical pitfalls. The anatomic knowledge required to avoid complications in surgery in this region is emphasized.
REGIONAL PATHOLOGY AND DIFFERENTIAL DIAGNOSIS Various disease processes that can involve the middle cranial fossa include pathology of the nerves, ganglia, dura, and bone. The most common tumors of the middle fossa are chordoma, chondrosarcoma, meningioma, trigeminal schwannoma, and osteosarcoma. Other lesions commonly operated on include cholesteatoma, cholesterol granuloma, squamous cell carcinoma, and lymphoma.
Chordoma Chordomas originate from the primitive midline notochord. They usually develop in the sacrococcygeal region but can manifest in the clivus region where they commonly spread throughout the skull base and into the middle cranial fossa. Meyers et al. reported extension into the middle fossa in 32.1% of cases (1). Although these tumors demonstrate aggressive local invasion, they rarely metastasize. The overall incidence of chordomas ranges from 0.1% to 0.7% of patients with intracranial tumors with occurrence in patients of all ages (2). Some studies suggest male predominance. Diagnosis is suggested by clinical history and examination as well as computed tomography and magnetic resonance imaging (MRI) (Fig. 1). Clinical symptoms depend on the degree of extension of the tumor to the regional neurovascular structures. On gross examination, chordomas are soft and gelatinous. Histological specimens demonstrate lobular patterns of uniform cells and vacuoles known as physaliphorous cells. Extracellular mucoid substance is present and necrosis is uncommon. A subset of chordomas, which is characterized by excess cartilaginous matrix, is known as chondroid chordomas. While some resemble chondrosarcomas, these entities demonstrate immunophenotypical patterns most consistent with chordomas. There is also an embryologic distinction— chordomas have an ectodermal origin and chondrosarcomas have a mesodermal origin. Chordomas, including the chondroid chordomas, stain positive for cytokeratin and epithelial membrane antigen, whereas chondrosarcomas do not (3,4). Difficult to cure via surgical resection, recurrence rates of skull base chordomas are often high. Survival of patients with untreated chordoma averages 28 months after onset of symptoms. Survival after surgery and/or radiation therapy ranges from 3.6 to 6.6 years. Recurrence usually occurs 2 to 3 years after initial treatment (2).
SURGICAL ANATOMY The floor of the middle fossa is formed by the greater sphenoid wing and the squamous and petrous parts of the temporal bone. The anterior and posterior borders are the sphenoid and petrous ridges, respectively. The medial part of the middle fossa is formed by the tuberculum sellae, pituitary fossa, both anterior and posterior clinoid processes, carotid sulcus, and dorsum sellae. The greater and lesser sphenoid wings form the lateral aspect. Several foramina, along with their related neurovascular structures, are intimately associated with the middle fossa. The optic canal and superior orbital fissure provide a passageway between the cranial base and the orbit. Foramina rotundum, ovale, and spinosum transmit the maxillary nerve, mandibular nerve, and middle meningeal artery, respectively. The anterior surface of the petrous bone, which forms the posterior boundary of the middle cranial fossa, has several important landmarks. Understanding these landmarks and their topographic relationship is essential for surgical planning in the middle cranial fossa. Meckel’s cave and the trigeminal ganglion are located medially near the apex of the petrous bone. Along the anterior border of the petrous bone are also Glasscock and Kawase triangles. The boundaries of Glasscock triangle form a line from foramen spinosum to the facial hiatus laterally, the greater superficial petrous nerve (GSPN) medially, and the mandibular nerve anteriorly (base). Kawase triangle is outlined by the GSPN laterally, the petrous ridge medially, and the arcuate eminence at the base. Although it is often mentioned that the petrous carotid artery lies within Glasscock triangle, it is more specifically beneath the GSPN. The superior semicircular canal lies beneath the arcuate eminence. The superior semicircular canal and the GSPN form an angle of approximately 120 367
368
Baaj et al.
Figure 2 T1-weighted brain MRI with contrast demonstrating strongly enhancing irregular bony lesion along the left cavernous sinus that extends into the left temporal fossa. Biopsy confirmed chondrosarcoma. Figure 1 Contrast-enhanced head CT showing suprasellar expansile, lytic bony lesion with parasellar extension. Biopsy confirmed chordoma.
Chondrosarcoma Chondrosarcomas of the skull base are rare, representing less than 0.1% of all head and neck tumors (5). The mean age of tumor presentation is the fourth and fifth decade of life without sex predilection. It is theorized that they develop from residual endochondral cartilage in the temporo–occipital junction, middle cranial fossa, sphenoethmoid area, anterior cranial fossa, and clivus. There are five histologic subtypes that predict behavior. The hyaline subtype, described as neoplastic chondrocytes, resides within lacunar spaces surrounded by a hyaline matrix. The myxoid subtype has chondrocytes surrounded by frothy mucinous matrix. Mesenchymal, dedifferentiated, and clear cell subtypes describe aggressive tumors with more anaplastic appearance. Treatment strategies include surgical debulking, complete resection, radiation therapy, and chemotherapy. Surgery invokes the principle of maximal safe resection (Fig. 2).
the sinus, especially those encasing the carotid artery, are best subjected to Simpson III or greater resection that entails removal of the extracavernous portion of the tumor. The intracavernous portion should be subjected to surveillance or radiosurgery.
Meningioma Meningiomas are extra-axial, slow-growing, benign tumors of the meninges. They are the most common primary intracranial tumors, and are usually located along the convexity, parasagittal region, and sphenoid ridge (Fig. 3). Sphenoid wing meningiomas are traditionally classified by Cushing and House as medial, middle, and lateral. For purposes of surgical planning, we group these lesions into clinoidal, sphenocavernous, and sphenoid wing types. Sphenoid wing meningiomas are converted to convexity meningiomas once the wing is removed, allowing for Simpson I resection. Although this is also possible for clinoidal meningiomas, opening of the optic canal for removal of en plaque canalicular extension is generally required. Simpson I resection of sphenocavernous meningiomas is limited to those that infiltrate the lateral wall without actual invasion into the sinus. The ones that invade
Figure 3 T1-weighted brain MRI with contrast showing suprasellar homogeneously enhancing lesion consistent with meningioma.
Chapter 24: Tumors of the Middle Cranial Fossa
369
Trigeminal Schwannoma Trigeminal schwannomas are rare, representing 0.07% to 0.36% of all intracranial tumors (6). Their incidence peaks in the fourth and fifth decades; they occur more often in women. Histologically, intracranial schwannomas demonstrate an intermediate organization of cells as compared with the classic Antoni A (well ordered arrays of elongated, spindle-shaped cells) and Antoni B (poorly organized large vacuolated cells) types. Trigeminal schwannomas are typically categorized into five groups as posterior fossa tumors, tumors of the Gasserian ganglion, dumbbell-shaped tumors, neurinomas of peripheral branches, and trigeminal schwannomas that involve multiple fossae. Small or asymptomatic lesions are often observed or undergo radiation. Surgical strategy should focus on radical resection of tumor with preservation of function. Specific approaches should account for the almost universal finding that trigeminal schwannomas have a Meckel cave component. Therefore, even with significant encroachment into the posterior fossa, our approach typically begins in the middle cranial fossa and opening of Meckel cave (Fig. 4).
Figure 5 T1-weighted brain MRI with contrast showing right petrous apex cholesterol granuloma.
Other Tumors In addition to metastatic disease, other rare entities must be included in the differential diagnosis of tumors of the middle cranial fossa, such as lymphoma (5) and osteosarcoma (9). More common lesions include cholesteatoma and cholesterol granuloma (7). Cholesteatoma is a lesion with a squamous epithelial–lined cavity that is filled with keratin and stratified squamous epithelial cells. Cholesterol
granuloma, commonly found at the petrous apex, develops as a result of hypoxia-induced hemoglobin degradation and accumulation of cholesterol crystals that cause a local inflammatory process (10) (Fig. 5).
CLINICAL ASSESSMENT Signs and symptoms of middle cranial fossa tumors can be either subtle or highly specific. In a series of patients with chordomas and chondrosarcomas, the most common symptoms were headache (65%), double vision (60.2%), and hoarseness/dysphagia (30.9%). The most common signs were palsies of the cranial nerves VI (47.2%), V (29.3%), and III (18.7%) (8). Clinical symptoms associated with trigeminal schwannomas depend on the location. About 50% of trigeminal schwannomas occur within the middle cranial fossa; their most common symptom is ipsilateral facial numbness. Facial pain similar to that caused by trigeminal neuralgia, is more common with a trigeminal ganglion tumor or lesion than with tumors of the trigeminal nerve root with the deficit usually involving all three divisions. Findings include diminished or absent corneal reflex and weakness in the muscles of mastication.
DIAGNOSTIC IMAGING
Figure 4 T1-weighted brain MRI with contrast showing right dumbbell schwannoma.
The principal imaging modalities in the evaluation of middle cranial fossa tumors are MRI and CT. Plain films and angiography may play an ancillary role. Nonenhanced CT scans provide detailed information on bony architectural changes associated with various tumors. Lytic bony destruction, particularly near the petroclival junction, is characteristic of chordomas and chondrosarcomas. In contrast, meningiomas are associated with hyperostosis and calcifications. While tumors in the middle cranial fossa may demonstrate variable intensity patterns on T1- and/or T2-weighted MR images, enhancement is usually prominent with gadolinium. Contour of the tumor, extent of invasiveness, and some tumor-specific signs may help differentiate these tumors on MRI scans. Chordomas and chondrosarcomas are hypo- to isointense on T1-weighted images, demonstrate
370
Baaj et al.
hyperintensity on T2-weighted images, and show heterogeneous enhancement with contrast (Fig. 1). Both of these tumors can have irregular contours and are locally invasive. Meningiomas are sharply demarcated lesions that frequently enhance homogeneously; a characteristic dural tail can often aid in making the diagnosis. The CT and MR characteristics of skull base lesions are well described. A few observations not always noted should be emphasized. For instance, cavernous sinus tumors that track along nerves are probably malignant and are unlikely to be benign meningiomas. On the other hand, encasement of the carotid artery is a typical feature of meningiomas. Bony lesions, such as chondrosarcoma, will instead displace the vascular structures.
PREOPERATIVE PREPARATION The first step in the preoperative preparation entails a thorough discussion of the risks and benefits with the patients and their families. This discussion should include specific potential complications of MCF tumor resection, such as cranial nerve palsies, particularly of the V and VII cranial nerves. A mandatory step in preoperative planning is a comprehensive neurologic examination with special attention to the cranial nerves. Corneal reflexes, oculomotricity, face sensation, and strength should be documented. If there is diminished hearing, a baseline audiogram should be obtained. Thin-cut CT slices through the floor of the middle cranial fossa or the petrous temporal bone can be obtained to better delineate the anatomy of the inner ear in relationship to tumor and bone destruction.
SURGICAL TECHNIQUE Exposure of the middle fossa is usually achieved via a standard temporal craniotomy. Because these tumors often extend beyond the middle cranial fossa, other surgical components can be added to the temporal craniotomy to expand exposure. A pterional craniotomy with or without resection of the orbital rim can extend the exposure forward. A zygomatic osteotomy can bring the exposure flush with or even through the middle fossa floor. Anterior petrosectomy can allow resection of tumors that extend deep into the petrous apex or into the posterior fossa.
Surface Landmarks Two surface landmarks, important for surgical exposures of the middle fossa, are the external auditory canal (EAC) and the root of the zygoma. The EAC is almost perfectly aligned with the IAC, both in the coronal and axial planes. The geniculate ganglion will also be found in the same coronal plane as the EAC and IAC. The root of the zygoma is the external landmark for the floor of the middle fossa. The root has both a vertically oriented component, which is lateral, and a horizontal component, which is medial and connects it to the squamous temporal bone. A coronal cut that runs through the center of the zygomatic root also runs through the foramen ovale. The depression that overlies the lesser wing of the sphenoid marks the Sylvian fissure and the anterior wall of the middle cranial fossa. The root of the zygoma marks the fossa floor near its junction with petrous bone. Similarly, a coronal cut that runs through the posterior aspect of the root of the zygoma will run through the foramen spinosum. These external landmarks and their reference to internal structure are important to properly position the craniotomy and to orient the sur-
geon during the extradural exposure of the middle fossa. Sufficient knowledge of surface landmarks allows the surgeon to translate external landmarks to internal space. Otherwise the craniotomy is incorrectly positioned to the target.
Selection of the Approach Following the building block concept to define a surgical approach, we start with a standard temporal craniotomy, which is the only stand-alone component. This approach is frequently sufficient for removal of parenchymal lesions. Skull base lesions often require additional components to expand the approach. A zygomatic osteotomy expands the approach inferiorly to the level of the middle cranial fossa. Each degree of bone resection decreases amount of brain retraction needed to gain the same visualization. The ability to create an up view is extremely useful in lesions that are medial to the temporal bone (e.g., cephalad extensions of petroclival meningiomas or basilar aneurysms). When a tumor extends to the anterior part of the middle fossa and involves the superior orbital fissure and lesser sphenoid wing, a pterional craniotomy is required as an additional step to expand the approach. Addition of a pterional craniotomy significantly shifts the exposure forward, thus allowing resection of the sphenoid wing, maxillary strut, and lateral bony wall of the superior orbital fissure. Finally, if the tumor is embedded in the petrous apex or extends from the middle into the posterior fossa, the anterior petrosectomy (Kawase approach) lowers the medial petrous apex down to the level of the IAC and creates a window between the two fossae. The exposure provided by an anterior petrosectomy is constant and predictable from preoperative imaging studies. The floor of the viewing trajectory through the anterior petrosectomy approach is the IAC and cisternal course of the cranial nerve VII–VIII complex. The scope of view in this approach becomes progressively constricted from top to bottom. The anterior petrosectomy cannot be expanded by the surgeon because it is limited by the IAC laterally and the inferior petrous sinus medially. The abducens nerve traverses the inferior petrous sinus. A tumor, however, can expand the anterior petrosectomy approach by naturally eroding the petrous apex. Such lesions can often be resected entirely through a middle fossa approach even when their caudal extent reaches as low as the jugular foramen. When approaches are planned for large tumors, all of the building blocks described can be combined to obtain a complete exposure of the middle fossa, petrous apex, and upper posterior fossa.
Positioning The position of the patient’s head for surgery is target specific. The classic position for lesions within the confines of the middle cranial fossa uses rotation of the head contralateral to the target. For example, with the sagittal sinus parallel to the floor, the patient’s head is tilted downward so that the zygoma is the highest point in the field. In young patients with a supple neck, this position can usually be achieved with an ipsilateral shoulder roll. If possible, the capability of the patient to sustain this head position for 5 to 10 minutes should be tested in an office setting. In older patients or those with neck stiffness, the park bench position is used. When other components (e.g., FTOZ or posterior petrosectomy) are added to the middle cranial fossa approach, head position is adjusted accordingly. Table rotation can accommodate some head position changes during surgery. On the occasion when a single position cannot accommodate the approaches necessary to define the target, we recommend staging the procedure (Fig. 6).
Chapter 24: Tumors of the Middle Cranial Fossa
371
of temporal craniotomies is based on our ability to envision the middle fossa on the surface of the bone as if the bone was transparent. The projection of the zygomatic root onto Macewen triangle defines the anterior face of the petrous bone. Closure and the cosmetic result are considered in both the skin incision and craniotomy. The incisions are designed to be hidden behind hairlines and to preserve wide-based vascular pedicles. If the zygomatic osteotomy is performed, it remains attached to the masseter muscle to preserve function when it is reattached.
Exposure of the Middle Fossa
Figure 6 An ideal position for MCF surgery, the sagittal sinus is parallel to the floor and the patient’s head tilts downward, making the zygoma the highest point in the field. Source: Courtesy of Mayfield Clinic.
FAQ: Do You Use a Lumbar Drain? We always place a lumbar drain in surgery in the middle cranial fossa because we believe that decreasing cranial CSF volume facilitates safe retraction of the brain. This is especially critical in the fossa where temporal lobe damage can be devastating. Drainage of CSF is in 20-mL aliquots as necessary; the drain is removed at the end of surgery.
Execution of the Approach Skin Incision After outlining the first craniotomy, the surgeon draws the simplest incision that will expose that craniotomy. For a small approach, a 5-cm straight incision over the center of the craniotomy often covers that exposure. As the approach expands forward, we shift toward a more conventional reverse question-mark incision. When the approach shifts posteriorly, we prefer a more classic neuro-otologic periauricular incision. When all of these approaches are needed, we combine these incisions.
The most common cause of poor exposure in the middle cranial fossa is inadequate dural elevation. The dura must be elevated aggressively from the face of the petrous bone posteriorly and from the fossa floor anteriorly to achieve sufficient exposure. The next impediment is the tethering of the dura at the foramen spinosum. Coagulation and sectioning of the middle meningeal artery are keys for adequate exposure. The dura in this region is best elevated from posterior to anterior direction to avoid picking up and tearing the GSPN. The foramen spinosum is identified by the middle meningeal artery at the dura and by following it to the foramen. Elevation of the dura continues medially to expose the true ridge of the petrous bone. Novices are often mistaken into exposing only as far as the anterior false ridge. These two ridges are created by the indentation created by the superior petrous sinus: the anterior ridge is false and the posterior is true. The superior petrous sinus must be elevated from the petrous ridge intact (Fig. 7).
FAQ: How Can I Be Certain That It Is the Foramen Spinosum Before Cutting the Structure That Runs Through It? The best way to confirm is to go back to the dura, find the middle meningeal artery, and follow it. Second, confirm that there is a larger foramen just anterior to it. Third (and not a good criteria), if you did not cauterize it, and you cut it and it does not bleed, it is probably not the foramen spinosum.
Craniotomy Two questions must be addressed up front in planning a craniotomy. First, do you intend to look medial to the temporal lobe (high basilar aneurysms, petroclival meningiomas)? If yes, a zygomatic osteotomy is performed to bring the temporalis muscle down off the temporal bone. Second, do you intend to look down? If yes, for lesions (e.g., meningiomas) that enter into the infratemporal fossa or angiofibromas that arise from the infratemporal fossa, the floor of the middle cranial fossa should be removed. Except for those cases, middle cranial fossa craniotomies should be flush with floor; this step is often critical but often not achieved. The temporal bone needs to be drilled until it is absolutely flat along the irregular fossa floor. Another principle, which maintains that the craniotomy should be centered on the target, requires understanding the superficial landmarks and their relationship to the lesion. The classic superficial landmark is the root of the zygoma and the EAC. Both are relative markers of the IAC. Our practice is to study a dry skull in preparation for a craniotomy. Development
Figure 7 Anatomy of the middle fossa with Glasscock and Kawase triangles delineated. Source: Courtesy of Mayfield Clinic.
372
Baaj et al.
FAQ: What Do I Do if I Get Totally Lost? When you are totally lost, go to the foramen rotundum. In the middle fossa, we call the foramen rotundum “home” because it represents a very reliable and easily identifiable landmark. When you dissect the dura anteriorly until bumping into the rise of the sphenoid wing, you will always see the maxillary division in the dural sleeve going into the foramen. Once the foramen rotundum is identified, the foramen ovale can easily be localized posteriorly and slightly lateral to it. The foramen spinosum will always be posterior and lateral to the foramen ovale. Only do this when you are lost because you will encounter irritating venous bleeding along the path.
FAQ: How Do I Find the GSPN? Bring your stimulator! Indeed, the first nerve that will be found during elevation of the middle fossa dura is the lesser superficial petrous nerve (LSPN) that runs lateral and parallel to the GSPN. The GSPN, which does not look like a nerve, exits the facial hiatus and imitates the periosteum that did not leave with the dura; it will stimulate. The LSPN conveys presynaptic parasympathetic fibers from the glossopharyngeal nerve (CN IX) to the otic ganglion where, after synapsing, it provides secretory fibers to the parotid gland. If the GSPN does not stimulate, that structure is the LSPN. Therefore, the LSPN will not stimulate the facial nerve. Conversely, stimulation of the GSPN will result in a positive contraction signal on neurophysiologic monitoring.
FAQ: Should the GSPN Be Cut to Prevent Traction Injury to the Facial Nerve? Although this was advocated in the past, we have completely abandoned the preventive section of the GSPN. Indeed, if the absence of lacrimation from the ipsilateral eye is only bothersome when isolated, it could prove to be a catastrophic complication if during the tumor resection, an injury to the V1 portion of the trigeminal nerve adds corneal numbness to an already dry eye. (Anatomic substrate: GSPN carries parasympathetics to the lacrimal gland.)
FAQ: What Retractor System Do You Need to Maintain an Optimal Exposure? Rather than the type of retractor, it is the position of the retractor blades that is important. Every effort should be made to place the blades behind the true ridge of the petrous bone rather than just on the temporal lobe dura.
Drilling of the Petrous Apex (Anterior Petrosectomy) The first step is to define the borders of Kawase triangle, including the true petrous ridge, arcuate eminence, mandibular division of the trigeminal nerve, and GSPN. With the intention to preserve the GSPN, all drilling is lateral to it. To maximize anterior petrosectomy, the lateral surface of the petrous ICA should be exposed. Exposure is accomplished by identifying the ICA at the site of dehiscence or drilling under irrigation just medial to the GSPN. Drilling can be move aggressive in the medial safe zone. This area lies in the medial Kawase triangle between the foramen ovale anteriorly and the arcuate eminence posteriorly. The bone in this area is thick; the IAC is the only vulnerable structure but is situated deep in the bone. In the anterior aspect of the safe zone, drilling exposes the posterior fossa dura, which serves as a road map to follow. Drilling stops when the inferior petrous sinus is encountered in the depth of the dissection. Although this structure can be sacrificed with significant labor, the abducens nerve penetrates the inferior petrous sinus
and will likely be injured. In the posterior aspect of the safe zone, drilling continues until dura is identified in the bone. That dura, which is the IAC, will serve as a roadmap to bone removal laterally. Medial-to-lateral exposure of the IAC is preferred because the nerves are more vulnerable laterally. A small arch of hard cortical bone is preserved in the angle between the GSPN and the arcuate eminence, which contains the cochlea. We estimate the location of the cochlea without specific identification and to date this has been effective in preserving it.
FAQ: How Do I Find the ICA If There Is No Dehiscence and Cannot be Seen On the Floor? Since a dehiscence occurs only in 15% of cases, this is common. Just start drilling medial to the GSPN. If you encounter muscle, this is the tensor tympani and the carotid artery just medial to it. In this procedure, the ICA is being exposed with the intention to maximize the petrosectomy and to not to manipulate the ICA in anyway.
FAQ: What Is the Best Way to Find the IAC? The first step is to predict its location. We do this by essentially bisecting the 120-degree angle created by the GSPN and arcuate eminence. As previously stated, we begin drilling medially in the safe zone until the dural sleeve of the IAC near the porus acousticus is identified; this dural sleeve is then followed from medial to lateral. We prefer this to the classic House technique in which the GSPN is followed to the geniculate and subsequently to the labyrinthine segment of the facial nerve. After identification of the IAC at its porus, it can be further exposed toward the fundus, while keeping in mind that its diameter decreases. At the level of Bill’s bar, the cochlea is located within a few millimeters anterior to the canal whereas the vestibule lays a few millimeters posterior to it. Determining the location of Bill bar with a right-angle nerve hook, rather than by direct visualization, will add a few millimeters of safety to prevent accidental opening of the basal turn of the cochlea. Although in laboratory dissection of the temporal bone, the cochlea is blue lined to ensure the maximal amount of drilling, this step is unnecessary in clinical practice. Contrary to the exposure of the ICA as the lateral limit of the exposure, no major advantage in viewing trajectory is gained with maximal drilling of the posterolateral corner of the Kawase quadrilateral to justify the risk of accidentally opening the basal turn with ensuing sensorineural hearing loss. After identification of the petrous ICA and the fundus of the IAC, drilling can proceed at a fast pace because no important structure is left in the petrous apex. It is important to leave a small shell of bone at the petrous ridge until the end of the drilling to keep the retractors in place.
Dural Opening If the tumor is confined in the middle fossa, a simple opening of the dura parallel to the middle fossa floor is sufficient to gain appropriate exposure. If an anterior petrosectomy was made, the goal of the dural opening is to eliminate any boundary between the middle and posterior fossae. To achieve this goal, the superior petrous sinus (SPS) and the tentorium will have to be cut; this delicate step in this dural opening is the clipping and cutting of this sinus. In its anterior third, the root of the trigeminal nerve crosses the petrous apex en route to the gasserian ganglion. The nerve is invariably in danger of being accidentally clipped because posterior fossa tumors
Chapter 24: Tumors of the Middle Cranial Fossa
373
The existence of a dissection plane between tumor and brainstem is determined by the presence of a subarachnoid plane on MRI scans; the softness of the tumor may occasionally be predicted based on MRI signal characteristics. Although, the softness or suckability of a tumor is generally the most critical factor in its resectability, it cannot be determined until the tumor, particularly meningiomas, is encountered during the operation. No specific technique is required for the tumor resection in this particular area. Tumor characteristics dictate which technique should be used (e.g., suction, ultrasonic aspiration, microscopic piece-meal resection, etc.) Knowledge of the anatomy and precise understanding of the preoperative studies are the best tools to preserve neurovascular structures. A nerve stimulator is helpful when the tumor approaches the cisternal or canalicular segment of the facial nerve.
Figure 8 Middle fossa contents after dural opening. Source: Courtesy of Mayfield Clinic.
always compress it against the tentorium and the SPS. As the dura and SPS are fairly tight at this step in the exposure, the only safe way to ensure preservation of the cranial nerve V is to section the SPS with a sharp blade millimeter by millimeter under good illumination and magnification; this technique enables identification of the fibers of the trigeminal nerve root that are compressed against the inferior aspect of the SPS wall. The sinus should be interrupted and cut between two medium-sized vascular ligaclips. Similarly, when sectioning the tentorium and reaching its free edge, identification of cranial nerve IV is mandatory before the final cut into the incisura is made (Fig. 8).
FAQ: How Do I Preserve the Trochlear Nerve? The trochlear nerve enters the tentorium fairly anteriorly and is therefore not directly threatened during dural opening. In the posterior part of its trajectory, it is protected by a layer of arachnoid that separates it from the tentorial edge. Although it should be identified before the final cut is made through the tentorial incisura, this can usually easily be done by either looking above or below the tentorium.
FAQ: What is All This Bleeding When I Cut the Tentorium? Venous lakes are present in the tentorium and should be coagulated before the cut is made. Nevertheless excessive bleeding is also encountered when the tentorium is cut too far anteriorly, near the origins of the superior and inferior petrous sinuses draining the cavernous sinus. When excessive bleeding is encountered, the cutting trajectory should be redirected more posteriorly. If the cutting trajectory is too posterior, the surgeon will follow the rise in the tentorium that goes behind the incisura and will never reach the incisura. These aberrant trajectories occur when the neurosurgeon fails to visualize the incisura before the incision is made.
Tumor Resection The most important concept in safe and effective tumor resection is the establishment of reasonable goals preoperatively, beginning with balancing the equation of a patient’s physiologic age compared with the predicted natural history of the disease. This assessment results in a set of decisions regarding partial versus total resection and the validity or nonvalidity of intentional neurologic deficits to achieve radical resection.
Closing FAQ: How Do You Close the Dura? Watertight dural closure is never achieved in middle fossa exposures when an anterior petrosectomy is also required. Important steps to avoid CSF leak are careful waxing of exposed air cells in particular around the IAC ostium, placement of a fat graft in the petrosectomy, and use of a dural sealant to keep the fat in place. As these patients are at risk for temporal lobe swelling, we discontinue the lumbar drain at the conclusion of surgery. The lumbar drain is probably unnecessary to prevent postoperative leak. Therefore, it is perceived to be harmful in maintaining multiple days of bed rest and producing headache.
POSTOPERATIVE CARE With the combination of the delicacy of the posterior temporal lobe and the small working place, temporal lobe contusion is not only a common postoperative finding but can also cause uncal herniation. Therefore, a postoperative CT is obtained on the day of the procedure. The severity of temporal lobe injury is exacerbated if the vein of Labb´e is injured. Clinical consequences are most severe on the dominant hemisphere. With the use of temporal lobe retraction and possibility of secondary temporal lobe seizures, we prescribe antiepileptic medications prophylactically to be used by our patients for 3 months. Postoperatively, the lumbar drain is discontinued at the end of surgery to prevent accidental over drainage. The patient typically requires a 24- to 48-hour stay in the ICU because of the potential for rapid deterioration. An immediate postoperative examination not only documents any new clinical findings, but can also detect a life-threatening hematoma. The efficacy of steroids in routine neurosurgical cases is controversial at best, and we typically rapidly taper the steroids after surgery.
COMPLICATION AVOIDANCE Not every neurologic deficit after surgery is regarded as a complication. Complications are events that are simultaneously undesirable, unintended, and uncommon; its substrate is generally lack of knowledge, poor planning, or poor execution. We have emphasized the anatomic knowledge required to avoid complications in surgery of the middle cranial fossa. We alluded to proper planning in discussing the adjustments
374
Baaj et al.
made in the surgical plan based on the patient’s physiology, tumor histology, and natural history of the disease. The rational for staging a procedure is based on limits of exposure for extensive tumors and the proportionality of specific complications to duration of surgery. Developing and maintaining an appropriate skill set for this kind of surgery depends on either high volume of rare cases or the availability of a cadaveric laboratory for study and practice. In addition to the common potential risks of craniotomies (i.e., bleeding, wound infection, etc.), several complications are specific to middle cranial fossa surgery. For example, the risk of temporal lobe contusion caused by excessive retraction can be reduced by CSF drainage via lumbar drain preoperatively or removal of the zygomatic arch. Another retraction-related injury is venous infarction, which is particularly deleterious if it involves the vein of Labb´e.
FOLLOW-UP AND REHABILITATION Follow-up should be based on the prediction of time to recurrence, which is based on understanding the natural history of the specific histology, degree of resection, and impact of surgical resection on that histology. These factors guide the strategy that implements surveillance or adjuvant treatment. In general, the rarity of many middle fossa tumors prompts us to seek the participation of our cancer center colleagues for evidence- and protocol-based management.
REFERENCES 1. Meyers SP, Hirsch WL Jr Curtin HD, et al. Chordomas of the skull base: MR features. AJNR. 1992;13:1627–1636. 2. Borba L, Colli BO, Al-Mefty O. Skull base chordomas. Neurosurgery Quarterly. 2001;11:124–139. 3. Crockard HA, Cheeseman A, Steel T, et al. A multidisciplinary team approach to skull base chondrosarcomas. J Neurosurg. 2001;95:184–189. 4. Crockard HA, Steel T, Plowman N, et al. A multidisciplinary team approach to skull base chordomas. J Neurosurg. 2001;95:175– 183. 5. Neff B, Sataloff RT, Storey L. Chondrosarcoma of the skull base. Laryngoscope. 2002;112:134–139. 6. Shrivastava RK, Sen C, Post KD. Trigeminal schwannoma. In: Winn HR, ed. Youman’s Neurological Surgery. 5th ed. Philadelphia, PA: W.B. Saunders, 2004:1343–1350. 7. Abdel Aziz KM, van Loveren HR. Primary lymphoma of Meckel’s cave mimicking trigeminal schwannoma: Case report. Neurosurgery. 1999;44(4):859–862. 8. Sekhar LN, Chanda A, Chandrasekar K, Wright D. Chondroma and chondrosarcoma. In: Winn HR, ed. Youman’s Neurological Surgery. 5th ed. Philadelphia, PA: W.B. Saunders, 2004:1283– 1294. 9. Alleyne CH, Theodore N, Spetzler RF, et al. Osteosarcoma of the temporal fossa with hemorrhagic presentation: Case report. Neurosurgery. 2000;47(2):447–451. 10. Pisaneschi MJ, Langer B. Congenital cholesteatoma and cholesterol granuloma of the temporal bone: Role of magnetic resonance imaging. Top Magn Reson Imaging. 2000;11(2):8–97.
25 Tumors of the Petrous Apex Ricardo Ramina, Maur´ıcio Coelho Neto, Yvens Barbosa Fernandes, Erasmo Barros da Silva, Jr., and Kristofer Luiz Fingerle Ramina
The inferior surface also has a foramen for the entry of the ICA. Medial to the jugular fossa is a depression, which is associated with the cochlear aqueduct (perilymphatic duct). The petrous bone articulates with the greater wing of the sphenoid anteriorly. The foramen lacerum is found between the apex of the petrous bone and the sphenoid bone and contains but does not transmit the ICA. The ICA penetrates the skull through the carotid canal of the temporal bone, and makes a curve medially to form the horizontal portion over the lacerum foramen penetrating the cavernous sinus. The segment of facial nerve mainly related (3–5 mm) to the petrous apex is the labyrinthine portion (3). The geniculate ganglion is anterior and medial to the arcuate eminence. In about 15% of the cases, the temporal bone is dehiscent over the geniculate ganglion (4). In children, the petrous apex is usually filled with fat becoming pneumatized with age like the mastoids. Embryologically its ossification is endochondral from mesenchymal tissue present in this region. The petrous apex may be pneumatic (20% of the population), sclerotic, or diploic. Pneumatization is relatively symmetric between right and left side and asymmetric pneumatization may be mistaken for tumor (5). There is no nervous or vascular structure within the petrous apex but it is surrounded by vital structures. These important vessels, nerves, and brain stem may be embedded or related to tumors arising in this region. Most surgical approaches require identification and dissection of these structures from the tumor capsule (6).
INTRODUCTION The petrous apex is located in the center of the skull base and is surrounded by critical structures making surgical removal of tumors from this region difficult. Several lesions may originate within the petrous apex producing bone erosion (primary lesions: e.g., chondrosarcomas, congenital cholesteatomas, cholesterol granulomas) or may arise from structures related to the petrous apex (secondary lesions: e.g., petroclival meningiomas, trigeminal schwannomas, chordomas). The surgical difficulties encountered in the treatment of petrous apex tumors are related to the involvement of the internal carotid artery (ICA) or basilar artery and their branches, brain retraction needed to approach the lesion, tumor extension to the brain stem and cavernous sinus, dissection and preservation of the vein of Labb´e, and reconstruction of the extensive surgical defects (after drilling of skull base bone) in order to prevent postoperative cerebrospinal fluid (CSF) leak and infection. The main challenge with this surgery is to achieve total tumor removal avoiding damage to cranial nerves, vessels, postoperative CSF leakage, and infection. Many surgical approaches have been used to resect these lesions. The following factors should be considered in choosing the most appropriate approach: nature of the lesion, anatomical location and extension, status of hearing and vestibular function, preoperative facial nerve function, presence of infection, and the experience of the surgical team. Neoplastic lesions produce symptoms through direct invasion and local compression “mass effect”.
REGIONAL PATHOLOGY AND DIFFERENTIAL DIAGNOSIS Embryologically, the petrous apex has a mesodermal origin and is formed by osseous and fat tissue. It is closely related to the clivus and the spheno-occipital and sphenopetrous synchondroses. These synchondroses have rests of notochordal and cartilaginous matrix. The temporal bone has pneumatized cells filled with mucosa of ectodermal origin. The cranial nerves related to this region also have an ectodermal origin. Lesions of the petrous apices may be infectious, inflammatory, neoplastic, or vascular (7–9). Tumors of the petrous apex may be classified into two groups: primary, originating within the petrous apex and tumors originating from neighboring structures with secondary involvement of the petrous apex. Petrous apex destruction is most frequently caused by a secondary process from either contiguous lesions or metastasis. Tables 1 and 2 show the most frequent lesions in both groups.
SURGICAL ANATOMY The temporal bone has four parts: petrous, squamous, tympanic, and mastoid. The petrous apex is like a pyramid with three surfaces between the middle and the posterior fossae (Fig. 1) (1). An anterior surface (temporal) makes up the medial part of the temporal bone, a posterior surface (posterior fossa), and an inferior surface (occipital). The anterosuperior portion of the petrous apex forms the middle fossa floor. The main anatomic structures related to this portion are the greater superficial petrosal nerve (GSPN) running posterior to the mandibular branch of the trigeminal nerve (2), the arcuate eminence, Eustachian tube, the ICA, and the gasserian and geniculate ganglia. The posterior surface of the petrous apex is the anterolateral wall of the posterior cranial fossa, and it extends medially from the posterior semicircular canal and the endolymphatic sac to the petroclinoid ligament and the canal for the abducens nerve (Dorello canal). This surface extends from the petro-occipital suture line inferiorly to the superior petrosal sinus superiorly. Inferiorly, the petrous pyramid is bounded by the jugular bulb and the inferior petrosal sinus.
Most Frequent Petrous Apex Lesions Chondrosarcomas Prognosis of petrous apex chondrosarcomas is related to histological grade. Grade I tumors have a 5-year survival rate of 375
376
Ramina et al.
Figure 1 (A) Lateral view of temporal bone. (B) Petrous apex—middle fossa and posterior fossa (arrows). (C) Petrous apex posterior view. (D) Anatomic structures of temporal bone and inner ear.
90%, grade II of 81%, and grade III of 43%. Facial nerve palsy, vertigo, diplopia (VI cranial nerve dysfunction), and pulsatile tinnitus are the most frequent complaints (10). Surgical approach depends on extension of the tumor. The middle fossa approach has been used in our clinic to remove petrous apex chondrosarcomas that extended medial to the ICA (Fig. 2). Radical resection is difficult in large and infiltrative tumors and postoperative radiosurgery is performed in high-grade tumors.
Table 1 Primary Petrous Apex Lesions Mesenchymal origin Chordoma Rhabdomyosarcoma Chondrosarcoma Neurofibroma Aneurysmal bone cyst Cholesteatoma Cholesterol granuloma Meningiomas Facial nerve schwannoma Trigeminal nerve schwannoma Ectodermic origin Epidermoid cyst Squamous cell carcinoma Mucocele Mesenchymal/ectodermic origin Multiple myeloma Lymphoma Metastasis
Table 2
Secondary Petrous Apex Lesions
Neoplastic Adenoid cystic carcinomas Juvenile angiofibromas Vestibular nerve schwannomas Nasopharyngeal carcinoma Chondrosarcoma Meningiomas Jugular foramen paragangliomas Chordoma Metastasis Non-neoplastic Epidermoid cyst Arachnoid cyst Fibrous dysplasia Intrapetrous carotid artery aneurysm Petrous apicitis Mucocele
Chordomas These tumors arise from notochordal remnants, are slow growing, and originate from the clivus and petrous apex (11). Recurrence after surgical removal is frequent and radiotherapy/radiosurgery (proton beam) is indicated as adjuvant therapy.
Chondromas They are rare, benign, petrous apex tumors that can be cured by radical resection (Fig. 3).
Chapter 25: Tumors of the Petrous Apex
Figure 2
377
(A) Preoperative MRI of large petrous apex chondrosarcoma. (B) Postoperative MRI after total removal.
l
Figure 3 (A) Preoperative MRI of a large petrous apex chondroma, (B) DSA showing displacement of the ICA, (C) surgical picture after total removal of the lesion (V3, third division of trigeminal nerve), and (D) postoperative CT scan.
378
Ramina et al.
Figure 4 Preoperative (A) and postoperative (B) MRI of a large petroclival meningioma.
Meningiomas Most meningiomas are benign and rarely originate from the petrous apex. Extension to this region from other sites (petroclival, Cerebellopontine angle, jugular foramen, and middle fossa) is more common. Surgical resection is the treatment of choice (Fig. 4).
Schwannomas Schwannomas affecting the petrous apex originate from the trigeminal and facial nerves. These tumors are slow-growing lesions and tend to be large at the time of clinical presentation. Complete resection is curative (Fig. 5).
Metastasis The most frequent metastases found in the petrous apex arise from the breast, lung, kidney, and gastrointestinal tract. Treatment and prognosis depend on the extent of the disease.
Cholesteatomas Petrous apex cholesteatomas may be congenital or acquired. Congenital lesions are rare and develop behind the tympanic membrane. Headaches, hearing loss, and facial nerve palsy are the most frequent symptoms. Large lesions may produce symptoms of other cranial nerves. Persistent otorrhea after previous mastoid surgery is an indication of an acquired cholesteatoma.
CLINICAL ASSESSMENT Petrous apex lesion may remain undetected for long periods. Symptoms may give some clues as to the histological diagnosis. Progressive and long-standing symptoms suggest benign tumors. Pain, multiple cranial nerve deficits, and short history are more frequently encountered with malignant lesions. Clinical symptoms are attributable to compression or infiltration of anatomic structures within or adjacent to the apex. The presenting signs and symptoms of petrous apex tumors may be specific (related to the structures of this region) or nonspecific and vague. Diagnosis is sometimes incidental on imaging
studies without related symptoms. Retroauricular pain and headache (retro-orbital and at the vertex) may occur with malignant or aggressive lesions from distortion of the dura. Facial pain, hypoesthesia, and paresthesia are observed with involvement of the trigeminal nerve at Meckel cave. Diplopia due to compression or invasion of the VI cranial nerve is observed in chondrosarcomas. The facial nerve may be affected anywhere along its course in the temporal bone. Tinnitus, vertigo, and hearing loss occur due Eustachian tube dysfunction, involvement of the bony labyrinth and the vestibulocochlear nerve, erosion of the ossicular chain, and compression of cerebellum and brain stem. Other cranial nerves from the II throughout the XII cranial nerve may be affected. Involvement of the ICA may produce pain (invasion of the adventitia), syncope, amaurosis fugax, and stroke.
DIAGNOSTIC IMAGING Evaluation of the temporal bone with standard radiography includes Towne’s and Stenver’s views and polytomography. An accurate diagnosis is, however, only possible with imaging studies such as CT scanning and MRI. CT scanning with contrast enhancement, thin slices, and 3-D reconstruction is useful to evaluate the extension of the lesion, erosion of the cranial base bone, and in planning the surgical approach. This diagnostic method may not accurately establish the diagnosis but is helpful in the differential diagnosis of lesions in this region (12) (Table 3). Tumors, fibrous dysplasia, multiple myeloma, and calcified chondrosarcomas are well demonstrated with CT scanning (13,14). MRI and MRA give more information (12,15) about the nature and extension of the lesion and involvement of other structures like vessels, nerves, and the brain stem (Table 4). Inflammatory diseases (e.g., petrositis, osteomyelitis) and epidermoid cyst may present characteristic findings on MRI. Petrous carotid and cavernous carotid aneurysms, as well as venous sinus variations, are well demonstrated with MR angiography and venography. Special sequences are helpful to visualize cranial nerves and relations of the tumor with the brain stem.
Chapter 25: Tumors of the Petrous Apex
Figure 5
Trigeminal nerve schwannomas. (A, B) Tumor in the temporal and infratemporal fossae. (C, D) Tumor in the posterior fossa.
Table 3 Petrous Apex Lesions on CT Scanning Lesion
Bone erosion
Eroded margin
Contralateral apex
Contrast enhancement
Cholesterol granuloma Cholesteatoma Petrous apicitis Effusion Bone marrow asymmetry Carotid aneurysm Neoplasia
+ + + − − + +
Smooth Smooth Irregular − − Smooth Variable
Highly pneumatized Often not pneumatized Variable Usually pneumatized Variable Variable Variable
− − − − − + +
Source: From Ref. 16.
379
380
Ramina et al.
Table 4 Petrous Apex Lesion on MRI Scanning Lesion
T1 images
T2 images
T1-gadolinium enhancement
Chondrosarcoma
↓(B)
↑↑↑(B)
Heterogenous
Chordoma
↓or↔(B)
↑↑↑↑(B)
Heterogenous
Cholesteatoma
↔(CSF)
↔(CSF)
No enhancement
Epidermoid cyst
↓↓(CSF)
No enhancement
Cholesterol granuloma
↑↑↑↑(CSF)
↑↑ or ↔(CSF) ↑↑↑↑(CSF)
Meningioma
↔(B)
↔(B)
Schwannoma
↔(B)
↑↑(B)
Carcinoma
↔(M)
↔(M)
Homogeneous Intense Homogeneous Intense Heterogeneous
Metastasis Plasmocytoma
↔(M) ↓ or ↔
↔(M) ↓ or ↓↓ or ↔
Heterogeneous Homogeneous Moderate
No enhancement
Tumor margins
Other characteristics
Irregular Invasive Irregular Invasive Regular Expansive Regular Expansive Regular Expansive Regular
Chondroid matrix
Regular Irregular Destructive Irregular Destructive Irregular
Bone “islands” Mmiddle line Prussak spaces ↑↑↑↑ Diffusion Confined Petrous apex Calcifications “Dura-tail” “Ice-cream cone sign”—cysts Extracranial extension Primary tumor Middle clivus
Abbreviations: ↓, hypointense; ↑, hyperintense; ↔, isointense; B, brain; M, muscle.
Digital angiography is performed when an aneurysm is suspected, when a balloon occlusion test is needed, and for preoperative embolization of paragangliomas and other highly vascular tumors.
SURGICAL APPROACHES
PREOPERATIVE PREPARATION
Translabyrinthine Approach
The preoperative preparation for treating these lesions involves evaluation of the patient’s clinical condition and associated comorbidities (Table 5). Many patients harboring petrous apex lesions are older than 60 years. The relative inaccessibility of this region often precludes a diagnostic biopsy and a careful analysis of the imaging studies and function of involved cranial nerves is needed to define the best form of treatment. Patients with associated infection (e.g., otitis) require preoperative antibiotic treatment. The quality of preoperative hearing is an important factor in deciding the surgical approach. Removal of the labyrinth should be avoided in patients with good hearing.
Table 5 Preoperative Preparation in Petrous Apex Tumors General evaluation
Specific evaluation
Preoperative anesthetic care
-Routine preoperative examination for a major surgery -Evaluation of comorbidities: heart, lung, kidneys, liver, diabetes, infection, and others. -History of thromboembolism, bleeding and use of drugs -CT with 3-D reconstruction -MRI, MRA -Angiography (embolization) -Audiometry, BAER, Facial and Trigeminal nerve functional testing -Adequate venous access for blood transfusion -Monitoring of invasive BP, CVP, O2 , CO2 -Monitoring of cranial nerves III, IV, VI, VII, VIII, IX, X, and XI (jugular foramen lesions) -Careful head rotation avoiding jugular vein and vertebral artery compression
Abbreviation: BAER, brain stem auditory evoked responses.
The choice of surgical approaches should consider the nature of the lesion, extension, involvement of the ICA, cavernous sinus, presence of facial nerve paralysis, preoperative hearing, vestibular function, and the presence of infection (1,17).
This approach is used in the treatment of petrous apex cholesteatomas, facial nerve schwannomas, and malignant lesions involving the temporal bone and the petrous apex (18). Resection of vestibular schwannomas and section of vestibular nerve may also be performed through this approach. In our department, however, all vestibular schwannomas are operated on using the retrosigmoid approach.
Surgical Technique The patient is operated on in the supine position under general anesthesia and contralateral head rotation. A retroauricular skin incision is made 2 cm posterior to the posteroauricular sulcus. The mastoid cortex is exposed after periosteal incisions and a complete mastoidectomy is performed. If infection is present (petrous apex cholesteatoma with secondary infection), marsupialization of the cavity and a wide meatoplasty using a skin flap are carried out to allow postoperative care of the cavity. In cases of facial nerve schwannomas or other tumors in patients with no useful hearing, a labyrinthectomy is performed to facilitate complete removal of the lesion. Management of the facial nerve depends on the preoperative facial nerve function, presence of infection, and if the proximal and distal stumps of the nerve can be identified. If the facial nerve is involved by the tumor but not infiltrated and its function is normal, the nerve is dissected from the lesion preserving the perineural tissues (neurolysis). If the nerve is infiltrated by the tumor and there is infection, reconstruction of the VII cranial nerve should be delayed. If there is no infection, the affected portion of the nerve is resected and reconstruction is carried out end-to-end or with nerve graft (sural nerve or great auricular nerve). In large defects when identification of the proximal stump is only possible at the brain stem, a sural graft is sutured with 10-0 nylon
Chapter 25: Tumors of the Petrous Apex
381
Figure 6 Petrous apex cholesteatoma (A, B). (C) Facial nerve reconstruction with sural nerve graft sutured in the stump at brain stem (arrows) and (D) distal stump in the parotid region (arrows).
or glued with fibrin glue at brain stem and at the stylomastoid foramen (Fig. 6). When a facial-to-facial anastomosis is not possible, a facial-to-hypoglossal nerve reconstruction is performed. After tumor removal and reconstruction of the facial nerve, the cavity is obliterated using muscle-periosteal flaps.
Middle Fossa Approach This approach exposes the petrous apex intradurally and extradurally (19). The second and third divisions of the trigeminal nerve, the gasserian ganglion, the petrosal portion of the ICA, and the meatal and petrosal portions of the facial nerve can be exposed through this surgical access. Petroclival meningiomas with their main portion in the middle fossa, facial and trigeminal schwannomas, small vestibular schwannomas, cholesteatomas, cholesterol granulomas, chondrosarcomas, teratomas, CSF fistulas, and facial nerve lesions at or medial to the genicular ganglion may be treated using this approach. Damage to the temporal lobe due to retraction or injury of draining veins is avoided using adequate anesthetic and microsurgical techniques.
Surgical Technique The patient is operated on under general anesthesia, with contralateral head rotation. A semicircular skin incision is cut beginning at the tragus extending anteriorly to the frontal region. A fascia temporalis flap is dissected exposing the temporalis muscle, zygomatic arch, and the upper portion of the external auditory canal. The temporal muscle is rotated down. If exposure of the infratemporal fossa is needed, the zygomatic arch is temporarily removed and in some selected cases opening of the glenoid fossa or resection of the mandibular condyle is performed. Temporal lobe retraction is reduced using this basal approach along with mannitol or lumbar CSF drainage. A craniotomy flap is cut exposing the basal portion of the temporal fossa from the middle portion of the zygomatic arch to the transverse sinus. The extradural approach to the petrous apex is carried out by elevating the dura mater of the middle fossa, exposing the middle meningeal artery, the GSPN, and second and third divisions of the trigeminal nerve. The middle meningeal artery is coagulated and cut. The GSPN is carefully dissected up to its exit in the facial nerve hiatus. The arcuate eminence (superior semicircular canal) is identified. This extradural
382
Figure 7
Ramina et al.
(A) Drawing showing skin incision and craniotomy. (B, C) surgical exposure of a facial nerve schwannoma (TU) in the geniculate ganglion region.
approach allows exposure of the petrous apex, internal auditory canal, the labyrinthine and tympanic portions of the facial nerve, and the petrosal and horizontal portions of the ICA by drilling of the bone medial to the Eustachian tube (Fig. 7). To expose the petrous apex region intradurally, a dura incision parallel to the sylvian fissure is made. The temporal lobe is carefully retracted avoiding damage to the draining veins, especially the veins of Labb´e complex. The free border of the tentorium is dissected; the IV cranial nerve is identified at the margin of the tentorium and the III cranial nerve anterior and medial to the fourth. Splitting the tentorial border allows exposure of the posterior fossa with V, VI, VII, and VIII cranial nerves (Fig. 8). The posterior communicating artery and its branches, the posterior cerebral and superior cerebellar arteries, are identified. All these anatomic structures may be displaced or embedded in the lesion and must be preserved. The posterior clinoid process and bone between the trigeminal nerve and the internal auditory meatus may be removed with a high-speed drill to enlarge the exposure medially. In cases of petroclival meningiomas, perforating branches from the basilar artery may be embedded in the tumor and radical removal may be impossible. After tumor removal, watertight dural closure is performed. All opened mastoid cells cells are closed with temporal muscle graft or
Figure 8 Anatomical dissection showing approach to the posterior fossa through the middle fossa. Cranial nerves III, V, VII and VIII. Abbreviation: ON, optic nerve.
Chapter 25: Tumors of the Petrous Apex
383
Figure 9 (A) Operative picture showing the asterium (arrow). (B) Retrosigmoid craniotomy. Abbreviations: TS, transverse sinus; SS, sigmoid sinus.
wax. The temporalis fascia flap is rotated to cover the dural opening and the craniotomy flap is replaced.
Presigmoid Approach
This approach is used to remove vestibular schwannomas, meningiomas of the cerebellopontine angle, and petroclival meningiomas with the main portion beneath the tentorium, as well as trigeminal schwannomas and epidermoid cysts (20). This approach permits wide exposure of the posterior fossa structures and the tentorial margin.
This approach is used for tumors with extension into the middle and posterior fossae (large petroclival meningiomas and trigeminal schwannomas are the most common examples) (21). In patients with good preoperative hearing, the labyrinthine block is not removed and the semicircular canals are not violated. If the patient has no hearing preoperatively, the labyrinthine block is removed to enlarge surgical exposure.
Surgical Technique
Surgical Technique
In our department, all patients are operated on in supine position with contralateral head rotation and elevation of the ipsilateral shoulder. This position has some advantages over the semi-sitting position: makes air embolism, arterial hypotension, and postoperative venous bleeding unlikely and is more comfortable for the surgeon in more prolonged surgeries. A retroauricular skin incision starting 5 cm above and 5 cm from the external auditory canal is performed. A small burr hole is drilled in the asterium region identifying the junction of the transverse and sigmoid sinuses. A 5-cm-diameter craniotomy is cut (Fig. 9). Its superior limit is the transverse sinus and the anterior limit is the sigmoid sinus. The mastoid emissary vein is coagulated and cut. If the sigmoid sinus is lacerated, it is repaired with 3-0 nylon stitches or a small muscle graft and fibrin glue. The opened mastoid cells are occluded with bone wax or muscle graft and fibrin glue to avoid postoperative CSF fistula. Under the surgical microscope, the dura mater is incised parallel to the sigmoid and transverse sinuses. The cerebellomedullary cistern is opened exposing the lower cranial, facial, cochlear, and trigeminal nerves, as well as the tentorium (Fig. 10). In cases of petroclival meningiomas, these nerves are dislocated posteriorly. Intracapsular tumor removal facilitates dissection of the cranial nerves from the tumor capsule. Tumor removal exposes the prepontine region and the petrous apex. Infiltrated dura mater and bone are removed to avoid recurrence. Tentorial opening and removal of the suprameatal recess and petrous apex allow better approach to the middle fossa. The dura mater is closed watertight and the craniotomy flap replaced.
The patient is placed in supine position with the head rotated to the opposite side. The skin incision starts in the temporal region and circumscribes the outer ear until the mastoid tip (Fig. 11). The flap is folded anteriorly exposing the temporalis muscle and the craniocervical fasciae. The temporalis muscle fascia is cut and dissected from the temporalis muscle. This fascia flap remains attached to the craniocervical fascia and both are rotated posteriorly with the sternocleidomastoid (SCM) muscle exposing the mastoid and retromastoid regions. A temporalis muscle flap is developed exposing the zygomatic arch and the middle fossa. These two flaps are used to cover the dura opening and the entire surgical field at the end of the procedure to prevent CSF fistula. A middle fossa approach is combined with a transmastoid approach. Extradural tumors (e.g., large petrous apex chondrosarcomas) are exposed after removal of the petrous apex medial to the arcuate eminence, carotid artery, and superiorly to the internal auditory canal. The dura mater is incised parallel to the middle fossa floor and presigmoid (Fig. 12). The superior petrosal sinus is ligated, sectioned, and the tentorium is opened exposing the petrous apex region. Temporal lobe retraction is minimal to avoid postoperative temporal lobe edema. The vein of Labb´e is preserved. The IV cranial nerve is dissected in the free edge of tentorium. The cranial nerves V, VII, and VIII are identified in the posterior fossa. Devascularization of feeder vessels is first performed in case of meningiomas before intracapsular tumor removal is started (Fig. 13). After reducing the size of the lesion, the tumor capsule is dissected from cranial nerves, vessels, and the brain stem. Some
Posterior Fossa (Retrosigmoid) Approach
384
Ramina et al.
A
B
Figure 10 (A) Retrosigmoid approach to a petroclival meningiomas exposing the tumor between cranial nerves. (B) Radical removal.
meningiomas invade the pia mater and total removal may result in damage to the brain stem. In these cases, a small portion of the capsule is not removed. The infiltrated dura mater and bone should be removed. All opened mastoid cells and pneumatized bone are covered with muscle graft and fibrin glue. The dura mater is closed watertight and the temporalis muscle flap is rotated over the dura mater incision. The fascia temporalis and craniocervical fascia SCM muscle flap are sutured over the craniotomy flap to cover the entire surgical field.
Occipito-Transmastoid-Cervical Approach In our department, this approach is used for resection of jugular foramen tumors with extension to the petrous apex and cervical region (e.g., paragangliomas, schwannomas, and meningiomas) (22–25).
Figure 12 Dura incision anterior to the sigmoid sinus (SS). Abbreviations: SPS, superior petrosal sinus; TS, transverse sinus.
Surgical Technique
Figure 11 Skin incision and drawing of craniotomy for presigmoid approach.
The patients are operated on in the supine position with the head turned 45 degrees to the opposite side. Care must be taken to not compress the opposite jugular vein. A “Cshaped” skin incision starting in the temporal region going down circumscribing the ear reaching the anterior border of the SCM muscle is performed (Fig. 14). The scalp is folded anteriorly. The external auditory canal is cut at the osteocartilaginous junction in cases with anterior extension of the lesion and when hearing is already lost. The temporalis muscle fascia is incised in the middle portion of the temporal region and dissected inferiorly until the temporal line exposing the temporalis muscle. About half of the temporalis muscle is incised, dissected, and turned down, at the end of surgery,
385
Chapter 25: Tumors of the Petrous Apex
Figure 13 Presigmoid approach to petroclival meningioma. Opening of tentorium and retraction of temporal lobe exposing the incisura with IV cranial nerve and the tumor.
to cover the dura, fill the entire mastoid cavity, and give a better cosmetic result. The cervical fascia is cut posterior to the external auditory canal, mastoid tip, and over the SCM. The insertion of the SCM is removed from the mastoid and a vascularized myofascial flap formed by the temporalis muscle fascia, the cervical fascia, and the SCM is turned posterior and inferiorly. At the end of surgery, this flap is secured with sutures in the temporalis, parotid, and cervical fasciae. It is turned back to cover the temporalis muscle flap and the entire surgical field. The next surgical step is neck dissection. The anterior border of the SCM is identified. The external jugular vein is ligated with suture/ligature and cut. The major vessels of the neck (common carotid, internal carotid, external carotid and its branches, and internal jugular vein) are dissected. The cranial nerves XII and VII are identified. The accessory nerve runs laterally to the jugular vein in the majority of cases and enters the upper portion of SCM. The vagus nerve and the sympathetic trunk run lateral and inferiorly to the common carotid artery. The vertebral artery is identified at the skull base after dissection of the lateral process of C1 and the superior and inferior oblique muscles. If the facial nerve is not infiltrated by the lesion, it is not
A
B
C
Figure 14 (A, B) Drawings showing the position and skin incision to approach a paraganglioma of jugular foramen. (C) Operative picture.
386
Ramina et al.
Surgical Technique
Figure 15 Anatomical specimen showing the facial nerve (arrows). In its bony canal after mastoidectomy, craniectomy, and neck dissection.
removed from its bony canal (Fig. 15). Grafts from sural or great auricular nerves are used to reconstruct the facial nerve when necessary. Radical mastoidectomy is performed with identification of the facial nerve canal, the labyrinth, mastoid antrum, ossicles, and the sinodural angle. The sigmoid sinus is exposed from the superior petrosal sinus to the jugular bulb. The dura anterior to the sigmoid sinus is dissected and the retrofacial mastoid cells are removed. Tumor within the ear, Eustachian tube, and mastoid cells, is removed. A 3-cm craniectomy limited by the transverse and sigmoid sinuses is performed. The jugular foramen is widely opened communicating its cervical portion with the cranial. The sigmoid sinus is ligated with two stitches below the superior petrosal sinus. The internal jugular vein is double ligated and cut and the extradural portion of the lesion is resected. The dura mater is incised in the medial wall of sigmoid sinus and the intradural portion of the tumor, the caudal cranial nerves at the brain stem, and cranial nerves VII and VIII are identified. In large lesions, compressing the brain stem, identification of cranial nerves is difficult. Step-by-step shrinkage of the tumor mass and intracapsular tumor removal permit identification of the cranial nerves at the brain stem and resection of the lesion. Primary dural closure or dural closure with fascia graft and fibrin glue is carried out. The posterior fossa is reconstructed with the vascularized flaps described above. Postoperative lumbar drainage is avoided. It is used only when the dural defect is large and cannot be closed in a watertight fashion.
Temporal Bone Resection This surgical procedure is indicated when a tumor involving the petrous bone can be totally resectable based on preoperative radiological studies. For benign tumors, this surgery may be curative (26). For low-grade malignancies, surgery and radiation is indicated. High-grade tumors are usually treated with adjuvant therapy and the recommendation for surgery depends on extension of the lesion. In our institution, high-grade malignancies, with tumor spread into the cavernous sinus, infratemporal fossa, upper cervical soft tissue, and medial to the petrous ICA, are not surgically treated because of the expected failure of this treatment modality and postoperative morbidity.
A “C-shaped” incision starting from the frontotemporal region to the neck is performed. Tumors infiltrating the skin or with infection may need a modified incision. The external auditory canal is transected at its osteocartilaginous junction, sutured and covered with a fascia flap to prevent CSF leak. The vessels (carotid artery and jugular vein) and cranial nerves are exposed in the neck. The facial nerve is dissected at the stylomastoid foramen. The temporal muscle is elevated from the middle fossa and a craniotomy is performed. The zygomatic arch, the squamous portion of the temporal bone, the condyle, and the neck of the mandible are cut with a drill. A mastoidectomy with skeletonization of the sigmoid and tranverse sinuses is carried out. A suboccipital craniotomy is performed. Extradural dissection of the temporal fossa exposes the GSPN and middle meningeal artery, which is coagulated and cut. The third division of trigeminal nerve is identified and the petrous carotid unroofed with diamond burrs. The opened Eustachian tube is packed with a piece of temporal muscle to avoid CSF fistula. If the sigmoid sinus is infiltrated, it is ligated with double ligatures and divided. The internal jugular vein is ligated and transected. The petrous apex and bone of the clivus are removed with a drill. The dura is incised in the middle fossa across the Meckel cave, superior petrosal sinus, tentorium, and posterior fossa behind the sigmoid sinus. The facial and vestibulocochlear nerves are transected at the internal auditory canal and the dura medial to these nerves is incised. Malignant tumors are resected “en bloc” (Fig. 16). Careful bipolar coagulation and Surgicel are used for hemostasis. Reconstruction of the facial nerve is performed with a sural nerve graft. The dural defect is closed with pericranium or bovine pericardium flap and fibrin glue. The temporalis muscle is rotated to cover the surgical field and sutured to the SCM muscle. Large defects are closed with rotation of trapezius or pectoralis flaps or with microsurgical free tissue transfer. Tarsorrhaphy is done to protect the eye and a lumbar CSF drain is inserted to reduce risk of CSF leak. Excessive CSF drainage should be avoided.
POSTOPERATIVE CARE The postoperative care includes all the general procedures after a major cranial surgery in an intensive care unit and specific measures related to involvement of structures of this region. Thromboembolism prophylaxis is initiated early, especially in elderly patients. Head CT or MRI is carried out on the day following surgery. Functional evaluation of the cranial nerves is performed as soon as possible. Facial nerve palsy, especially when associated with palsy of the first trigeminal nerve division, requires care of the eye to avoid trauma to the cornea. When the facial nerve is sectioned, it is preferentially reconstructed with an end-to-end suture or with a graft from the great auricular nerve or from the sural nerve. An anastomosis between the VII and XII cranial nerves is usually performed 2 weeks after surgery if a VII–VII reconstruction was not possible. A tarsorrhaphy is carried out as soon as possible in cases of complete facial nerve lesion. Postoperative dysfunction of cranial nerves IX and X causing dysphagia with bronchial aspiration and pneumonia may be a severe and even fatal complication. Patients with involvement of these nerves should be extubated only when they are awake and after careful evaluation of the function of the caudal cranial nerves. Tracheostomy and gastrostomy
Chapter 25: Tumors of the Petrous Apex
387
Figure 16 (A) CT scan showing a temporal bone carcinoma (arrows). (B) Skin incision. (C) Surgical picture after resection of temporal bone. (D) Facial nerve reconstruction with sural nerve graft.
are performed without any delay in patients with severe dysphagia and vocal cord palsy. In moderate cases of caudal cranial nerve dysfunction, the grade of reflux is evaluated with endoscopy and video fluoroscopy. These cases are usually managed with a nasogastric tube and phonorehabilitation. Procedures for vocal cord medialization may be helpful when these measures fail.
COMPLICATIONS AND AVOIDANCE CSF fistula is a common complication after intradural surgeries involving the mastoid and the temporal bone (Table 6). The described myofascial vascularized flaps were developed in our department to close the large surgical defect, thus avoiding CSF fistula. The cosmetic results using this reconstruction technique are excellent. Postoperative infection may
Table 6
occur due to intraoperative contamination or due to previous ear and mastoid infections. Careful otoscopy and evaluation of mastoid cells and paranasal sinuses is mandatory to diagnose infection in these regions. Damage to the greater petrosal nerve will produce “dry eye”. The petrous portion of the ICA may be dehiscent and careful drilling of the petrous apex bone is needed to avoid damage of this vessel. Drilling the petrous bone may damage the arcuate eminence (superior semicircular canal), the geniculate ganglion, and the cochlea anterosuperiorly. Damage to the cranial nerves within the jugular foramen may occur in tumors of the posterior portion of the temporal bone. Excessive temporal lobe retraction will produce venous compression, postoperative edema that may cause hemiplegia, aphasia, and even death. This complication may be avoided by using a basal approach with drilling of the temporal bone, good anesthesia, adequate
CSF Fistula in Temporal Bone Surgery
Fistula types
Prevention and management
Through mastoid cells
Prevention: Watertight dura mater closure. All opened pneumatized cells are closed with muscle graft, fibrin glue, and bone wax. Management: Lumbar drainage, acetazolamide for 3 days. In case of failure reoperation. Prevention: Watertight dura mater closure. Vascularized myofascial flap rotation. Management: Compressive dressing, lumbar drainage, and reoperation in case of failure.
Subcutaneous
388
Ramina et al.
microsurgical technique, and in some cases lumbar drainage for brain relaxation. Identification and dissection of the vein of Labb´e is very important to avoid venous infarction of the temporal lobe. Postoperative complication due to damage of cranial nerves V, VII, IX, and X was described above.
REFERENCES 1. Ramina R, Maninglia J, Barrionuevo CE. Surgical excision of petrous apex lesions. In: Sekhar LN, Janecka IP, eds. Surgery of Cranial Base Tumors. New York: Raven Press Ltd, 1993:291– 305. 2. Kruger L, Young RF. Specialized features of the trigeminal nerve and its central connections. In: Samii M, Jannetta PJ, eds. The Cranial Nerves. Berlin: Springer, 1981: 273–301. 3. Proctor B. Canals of the temporal bone. In: Surgical Anatomy of the Ear and Temporal Bone. New York: Thieme Medical Pub., 1989:89–128. 4. Rhoton AL Jr, Hall GM. Absence of bone over the geniculate ganglion. J Neurosurg. 1968;28:48–53. 5. Roland MS, Meyerhoff WL, Judge CO, et al. Asymmetric pneumatization of the petrous apex. Otolaryngol Head Neck Surg. 1990;103:80–88. 6. Fournier HD, Mercier P, Velut S, et al. Surgical anatomy and dissection of the petrous and peripetrous area. Anatomic basis of the lateral approaches to the skull base. Surg Radiol Anat. 1994;16(2):143–148. 7. Curtin HD, Som PM. The petrous apex. Otolaryngol Clin North Am. 1995;28(3):473–496. 8. Franklin DJ, Jenkins HA, Horowitz BL, et al. Management of petrous apex lesions. Arch Otolaryngol Head Neck Surg. 1989;115:1121–1125. 9. Muckle RP, De la Cruz A, Lo WM. Petrous apex lesions. Am J Otol. 1998;19(2):219–225. 10. Kveton JF, Brackmann DE, Glasscock ME, et al. Chondrosarcoma of the skull base. Otolaryngol Head Neck Surg. 1986;94:23–32. 11. Tzortzidis F, Elahi F, Wright D, et al. Patient outcome at long-term follow-up after aggressive microsurgical resection of cranial base chordomas. Neurosurgery. 2006;59(2):230–237.
12. Greess H, Baum U, Romer W, et al. CT and MRI of the petrous bone. HNO. 2002;50(10):906–919. 13. Brown RV, Sage MR, Brophy BP. CT and MR findings in patients with chordomas of the petrous apex. AJNR. 1990;11:121–124. 14. Hirsch WL Jr, Curtin HD. Imaging of the lateral skull base. In: Brackmann DE, Jackler RK, eds. Neurootology. St. Louis, MO: Mosby, 1994:303–340. 15. Struffert T, Grunwaqld IQ, Papanagiotou P, et al. Imaging of the temporal bone. An overview. Radiologe. 2005;45(9):816–827. 16. Jackler RK, Parker D. The radiographic differential diagnosis of petrous apex lesions. Am J Otol. 1992;13:561–574. 17. Meneses MS, Moreira AL, Bordignon KC, et al. Surgical approaches to the petrous apex:Distances and relations with cranial morphology. Skull Base. 2004;14(1):9–20. 18. Brackmann DE, Toh EH. Surgical management of petrous apex cholesterol granulomas. Otol Neurotol. 2002;23(4):529–533. 19. Youssef S, Kim EY, Aziz KM, et al. The subtemporal intradural approach to dumbbell-shaped trigeminal schwannomas: Cadaveric prosection. Neurosurgery. 2006;59(4 suppl):270–278. 20. Samii M, Tatagiba M, Carvalho GA. Retrosigmoid intradural suprameatal approach to Meckel’s cave and the middle fossa: Surgical technique and outcome. J Neurosurg. 2000;92(2):235– 241. 21. Samii M, Tatagiba M. Experience with 36 cases of petroclival meningiomas. Acta Neurochir (Wien). 1992;118(1–2):27–32. 22. Ramina R, Maniglia JJ, Fernandes YB, et al. Jugular Foramen tumors. Diagnosis and treatment. Neurosurg Focus. 2004;17(2):E5. 23. Ramina R, Maniglia JJ, Fernandes YB, et al. Tumors of the jugular foramen: Diagnosis and management. Neurosurgery. 2005;57(1 suppl):59–68. 24. Ramina R, Maniglia JJ, Paschoal JR, et al. Reconstruction of the cranial base in surgery for jugular foramen tumors. Neurosurgery. 2005;56(2 suppl):337–343. 25. Ramina R, Coelho Neto MC, Fernandes YB, et al. Meningiomas of the jugular foramen. Neurosurg Rev. 2006;29(1):55–60. 26. Pomeranz S, Sekhar LN, Janecka IP, et al. Classification, technique and results of surgical resection of petrous bone tumors. In: Sekhar LN, Janecka IP, eds. Surgery of Cranial Base Tumors. New York:Raven Press Ltd, 1992:317–335. 27. Glasscock ME III, Woods CI III, Poe DS, et al. Petrous apex cholesteatoma. Oto Clin North Am. 1989;22(5):981–1002.
26 Tumors of the Cerebellopontine Angle Bryan C. Oh, Daniel J. Hoh, and Steven L. Giannotta
facial and vestibulocochlear nerves arise from the brainstem near the lateral end of the pontomedullary sulcus, anterosuperior to the choroid plexus protruding from the foramen of Luschka (1). In most cases, the AICA passes below the facial and vestibulocochlear nerves as it encircles the brainstem. In this situation, the tumor would displace the artery inferiorly. The labyrinthine, recurrent perforating, and subarcuate branches arise from the AICA near the facial and vestibulocochlear nerves and are frequently stretched around a CPA tumor (1). The veins on the side of the brainstem that have a predictable relationship to the facial and vestibulocochlear nerves are the vein of the pontomedullary sulcus, the veins of the cerebellomedullary fissure, middle cerebellar peduncle, and cerebellopontine fissure (7). Identification of any of these veins during the tumor removal simplifies identification of the site of junction of the facial and vestibulocochlear nerves with the brainstem.
SURGICAL ANATOMY The anatomy of the cerebellopontine angle (CPA) was described in great detail several years ago by Rhoton (1). Briefly, the CPA cistern is bounded laterally by the petrous face, medially by the pons, and superiorly by the tentorium cerebelli (2). The cerebellopontine fissure opens medially and has superior and inferior limbs that meet at a lateral apex. Cranial nerves IV through XI are located near or within the angular space between the two limbs commonly referred to as the CPA (1). The superior cerebellar artery and anteroinferior cerebellar artery (AICA) both arise from the basilar artery and course through the CPA cistern (3). Veins from the pons, middle cerebellar peduncle, and cerebellopontine fissure unite near the trigeminal nerve and form the superior petrous veins (4). The trochlear and trigeminal nerves are located near the fissure’s superior limb, whereas the glossopharyngeal, vagus, and accessory nerves are located near the inferior limb. The abducens nerve is located near the base of the fissure, along a line connecting the anterior ends of the superior and inferior limbs (1). Although the facial and vestibulocochlear nerves may appear to pass as a single bundle from the pontomedullary junction to the internal auditory meatus, they are separate. The superior and inferior vestibular nerves lie posteriorly and superiorly, and the cochlear nerve posteriorly and inferiorly (Fig. 1). A shallow groove marks the boundary between them. The facial nerve lies anteriorly and slightly superiorly (2). The labyrinthine artery (and occasionally the main trunk of AICA) usually lies between the facial and vestibular nerves. However, specific anatomic relationships are often distorted as tumor enlarges (Fig. 2). As will later be discussed, acoustic neuromas, or vestibular schwannomas are by far the most common tumors that develop within the CPA. Discussion from this point will thus focus on surgical anatomy of acoustic neuromas. Within the lateral portion of the internal auditory meatus are the facial, cochlear, and inferior and superior vestibular nerves. The lateral portion of the meatus is divided into a superior and an inferior portion by a horizontal ridge called the transverse or falciform crest. The facial and the superior vestibular nerves are superior to the crest. The facial nerve is anterior to the superior vestibular nerve and is separated from it at the lateral end of the meatus by a vertical ridge of bone, called the vertical crest (Bill bar, named after William House) (5). The cochlear and inferior vestibular nerves run below the transverse crest with the cochlear nerve located anteriorly. Identification of Bill bar will therefore allow delineation of the nerves anterior to it (facial and cochlear) and the nerves posterior to it (superior and inferior vestibular). A consistent set of neural, arterial, and venous relationships at the brainstem facilitates the identification of the nerves on the medial side of an acoustic neuroma (6). The
REGIONAL PATHOLOGY AND DIFFERENTIAL DIAGNOSIS Tumors in the CPA cistern can arise from the brain, temporal bone, or subarachnoid space and its contents. A majority of CPA masses are located primarily in the cistern. Others arise within the internal auditory canal (IAC) itself, and some originate in the adjacent brain or skull and extend secondarily into the CPA (3). Table 1 provides a differential diagnosis for a mass within the CPA cistern and lists the frequency at which each of the tumors occurs. It should be noted that normal structures and anatomic variants may be mistaken for a CPA mass. These include the cerebellar flocculus, choroid plexus at the foramen of Luschka, a high jugular bulb, and prominent jugular tubercles (3). An acoustic neuroma, or vestibular schwannoma is by far the most common CPA mass, accounting for 75% of all lesions in this location (3). Acoustic neuromas account for 6% of all intracranial tumors. Meningioma is the next most common tumor in the CPA. Nakamura et al. recently studied the location of CPA meningiomas with respect to the IAC in 270 patients (8). Among these patients, 32.9% of the tumors originated at the petrous ridge anterior to the IAC, 22.2% showed involvement of the IAC, 20.2% were located superior to the IAC, 11.8% were inferior to the IAC, and 12.9% were posterior to the IAC, originating between the IAC and the sigmoid sinus. In a study of meningiomas of the CPA, Voss et al. reported that the most common site of dural origin was the petrous ridge [anterior to the IAC (26%), posterior (21%), superior (18%), and inferior (16%)]. Less common sites of dural origin included the tentorium (31%), the clivus (15%), the IAC (10%), and the jugular foramen (8%) (9). The site of dural origin determines the direction of displacement of the facial/vestibulocochlear nerve bundle. Epidermoid tumors 389
390
Oh et al. Table 1 Specific Tumors of the CPA and the Percentage of all CPA Tumors That They Comprise Type of mass Acoustic neuroma Meningioma Epidermoid Other cranial nerve schwannoma Metastasis Paraganglioma Ependymoma, choroid plexus papilloma, lipoma, chordoma
Anterior
Vestibular division of VIII is superior
Cranial Cochlear division of VIII is inferior
Percentage of all CPA tumors 75 8–10 5 2–5 1–2 1–2 1
or choroid plexus papillomas can also extend into the CPA. Finally, rare tumors such as chordomas, paragangliomas (11), and lipomas (12) can be present in the CPA. It is also important to note that 2% to 5% of CPA masses are not tumors. Vascular lesions such as vertebrobasilar dolichoectasia and aneurysms account for the majority of nontumor CPA masses. Although extremely rare, arachnoid cysts can also be found in the CPA.
CLINICAL ASSESSMENT Posterior
Figure 1 This is an intraoperative photograph demonstrating position of the inferior and superior vestibular nerves with respect to the cochlear nerve. The superior and inferior vestibular nerves lie posteriorly and superiorly, and the cochlear nerve posteriorly and inferiorly.
and schwannomas of other cranial nerves are the third and fourth most common tumors of the CPA, respectively (3). Metastases account for 1% to 2% of CPA masses. In their series of 14 patients with CPA metastases, Yuh et al. reported that 100% of patients had multiple cranial nerve involvement. Additionally, most patients had bilateral cranial nerve and leptomeningeal involvement (10). A variety of masses that originate in the fourth ventricle such as ependymomas
Aterioinferior cerebellar artery
Tumor
Trigeminal nerve
Cerebellum
Figure 2 This is an intraoperative photograph demonstrating the point that a tumor often distorts standard anatomical relationships when it enlarges. Specifically, this image is from a retrosigmoid approach for an acoustic neuroma.
Patients usually present with symptoms that are closely correlated with tumor size. In a series of 131 patients with acoustic neuromas, Harner and Laws noted that the three most common presenting symptoms are hearing loss, tinnitus, and dysequilibrium, respectively (13). Larger tumors were found to cause facial numbness, weakness, twitching, or possibly brainstem symptoms. Rarely, tumors were large enough to produce hydrocephalus. Nakamura et al. recently reported a small series of meningiomas limited to the IAC and noted that tinnitus, hearing loss, and vertigo or dizziness were the top three symptoms, respectively (14). Finally, patients with epidermoids also commonly present with trigeminal neuralgia or hemifacial spasm (15,16). Patients with CPA epidermoids can also present with ataxia, nystagmus, and lower cranial nerve dysfunction (17). With regard to clinical signs, 66% of patients had no abnormal physical finding except for hearing loss in the acoustic neuroma series of Harner and Laws (13). Since hearing loss is sensorineural, Weber’s test will lateralize to the uninvolved side. If enough hearing is preserved, Rinne test will be normal (air conduction > bone conduction). Facial nerve function should be graded on the House and Brackmann scale (18). Excluding hearing loss, the next three most common abnormal signs in patients with acoustic neuromas are abnormal corneal reflex, nystagmus, and facial hypoesthesia (13). When evaluating any patient who has a mass lesion in the CPA, a clinician should carefully examine the patient for these physical signs. A pure tone audiogram (PTA) is often the first screening test for patients when the diagnosis of acoustic neuroma is suspected. High frequency hearing loss is the most common abnormality seen on pure tone audiometry (19). However, this is also the most common type of hearing loss with age and from noise exposure. In general, a hearing difference from one ear to the other of greater than 10–15 dB on an audiogram without a good explanation should be investigated further. Johnson found that the likelihood of abnormal audiometry correlates with tumor size (19). Brainstem auditory evoked responses provide another preoperative measure in the diagnosis of sensorineural hearing loss. During this examination, a short duration pulse stimulus is applied to the
Chapter 26: Tumors of the Cerebellopontine Angle
ear, thereby evoking a number of peaks corresponding to defined structures along the hearing pathway. Typical findings associated with acoustic tumors include poor replicability, an abnormal interaural wave V (lateral lemniscus) latency difference, increased latency in the peaks associated with the superior olivary complex and the inferior colliculus as compared to the cochlear nerve, and increased absolute latency measurements. Another important aspect of the clinical assessment is speech discrimination, which is not simply related to the degree of pure tone hearing loss. Some patients may have exceptionally poor speech discrimination despite near normal pure tone audiometry (2). In a series of 425 patients with acoustic neuromas, only 20% of patients demonstrated good speech discrimination (19). However, it should be noted that this series is from 1977 and thus predates modern diagnostic imaging techniques such as magnetic resonance imaging (MRI). Patients from this series probably were diagnosed later and likely had poorer speech discrimination at time of diagnosis than patients today. Speech discrimination score is an integral part of the clinical decision-making process when considering hearing preservation procedures. When hearing in the contralateral ear is normal, residual hearing on the operated side is socially useful only if speech discrimination is good and the PTA of the normal side is within 30 dB (2).
DIAGNOSTIC IMAGING Over the past few decades, there has been a tremendous improvement in both the quality and sensitivity in diagnostic imaging. Imaging modalities have evolved through plain films, cisternography, angiography, computed tomography (CT), and MRI. Currently, MRI is the imaging investigation of choice when evaluating and following tumors of the CPA. The increased sensitivity for detecting CPA tumors is affecting the pattern of presentation and will likely influence management strategies and outcomes (2).
Computed Tomography Prior to MRI, CT was the primary imaging modality of choice for diagnosis of acoustic neuromas and other CPA tumors. Contrast-enhanced CT scans, with 5 mm axial slices through the cranial base, were used to detect all but the smallest soft tissue masses within the CPA (20). The classic CT appearance of an acoustic neuroma is an iso- or hypodense lesion centered upon the internal auditory meatus, with homogeneous enhancement after intravenous contrast. CPA meningiomas may have similar appearances but are usually hyperdense prior to contrast injection and commonly are placed asymmetrically in relation to the porus acousticus (2). Additionally, meningiomas generally have a flat, broad-based attachment to the petrous bone, whereas the angle between the tumor and the petrous bone should be acute in patients with acoustic neuroma (21). Dural edge enhancement adjacent to the main tumor is also highly suggestive of meningioma (22), as is the presence of calcification, which will be evident in roughly 25% of cases (23). Macroscopic calcification that is discernable on CT scan in acoustic neuromas is exceedingly rare (24). CT is the best imaging modality available for delineating bony anatomy. Details of bony anatomy can be important for several reasons. First, it may aid in establishing a diagnosis. This is often true for small tumors that show little enhancement, but that cause early expansion of the internal auditory meatus. Only rarely is the meatus enlarged
391
Figure 3 An axial T1WI after intravenous gadolinium administration from a patient with a left acoustic neuroma is shown. Acoustic neuromas generally readily enhance with contrast administration.
by a meningioma. However, erosion of the temporal bone by cholesteatoma, facial nerve schwannoma, or metastasis may be evident and assist in establishing a diagnosis (2). Additionally, CT can provide useful information for the surgeon. The size and location of perimeatal and labyrinthine air cells can be noted. If a high resolution CT scan is performed, the anatomic relationship between the semicircular canals, the vestibule, and the internal auditory meatus can be appreciated as can the position of the jugular bulb. These relationships could influence the choice of surgical approach. Bony erosion by tumor in the region of the jugular bulb may also be apparent (2).
Magnetic Resonance Imaging MRI is now the imaging modality of choice for diagnosis and following tumors of the CPA. Most acoustic neuromas are hypo- or isointense on T1-weighted images (T1WI) when compared to normal brain parenchyma. These tumors usually enhance markedly after administration of intravenous gadolinium. An axial T1WI after intravenous gadolinium administration from a patient with a left acoustic neuroma is shown in Figure 3. On T2-weighted images (T2WI), acoustic neuromas are usually hyperintense when compared to normal brain. Large tumors may have a fair amount of cystic degeneration (Fig. 4). The salient advantages of MRI over CT
Figure 4 Large acoustic neuromas can also have a fair amount of cystic degeneration. This is a contrast-enhanced axial T1WI from a patient with an acoustic neuroma.
392
Oh et al.
Figure 5 This contrast-enhanced axial T1WI shows a left CPA mass with a broad dural base, suggestive of meningioma. Note the significant extension of the tumor into the internal acoustic canal. Diagnosis was pathologically confirmed at the time of surgery.
are superior resolution, lack of beam-hardening artifact, ability to image tumor in multiple planes, and ability to identify certain vascular structures and possible vascular displacement or encasement (2). MRI has also been used to characterize the natural history of acoustic neuromas. In a series of patients who were treated conservatively and followed with serial scans, tumor growth rate was 0.91 mm per year (25). Nonsurgical tumors did not grow or regressed in 42%. Overall postoperative growth rate for surgical subtotal resection
patients was 0.35 mm per year. Surgical tumors did not grow or regressed after subtotal resection of acoustic neuroma in 68.5% of patients (25). Meningiomas may also appear hypo- or isointense on T1WI when compared with brain. They also strongly enhance with intravenous gadolinium administration. Compared with brain parenchyma, they are usually isointense on T2WI. A broad dural base or dural “tail” may help to distinguish meningiomas from other tumors (3). An example of a contrast-enhanced T1WI from a patient with a CPA meningioma is shown (Fig. 5). Hyperostosis may also be appreciated using newer MRI techniques, although CT is still preferable for observing this phenomenon. The most common location of intracranial epidermoids is the CPA (26). Epidermoids, which appear isotense when compared to CSF on T1WI, rarely enhance with intravenous gadolinium (Fig. 6). They may be iso- or hyperintense when compared to CSF on T2WI, and they very rarely demonstrate enhancement or calcification. Liu et al. evaluated patients with CPA epidermoids with more modern MRI techniques (26). Diffusion-weighted imaging depicted all lesions as strongly hyperintense relative to CSF and brain tissue. FLAIR (fast fluid-attenuated inversion recovery) sequences depicted the lesions with mixed signal intensities but with poor-to-medium demarcation. MR cisternography depicted the lesions as hypointense to CSF and clearly showed the anatomic relation to neighboring nerves and vessels. Finally, metastases in the CPA are usually isointense compared to brain parenchyma on both T1WI and T2WI. Lesions usually demonstrate a moderate amount of enhancement, and calcifications are exceedingly rare (3). Coexisting
Figure 6 Contrast-enhanced axial T1WI from a patient with a left CPA epidermoid (top panel). Note the similarity in signal intensity to cerebrospinal fluid on T2WI (bottom left panel), however, the lesion demonstrates evidence of restricted diffusion on DWI (bottom, middle panel). Source: Images courtesy of Dr. Hervey Segall, Department of Neuroradiology, University of Southern California, Los Angeles, CA.
Chapter 26: Tumors of the Cerebellopontine Angle Table 2
393
Comparative Imaging Findings of CPA Masses
Type of mass
T1 compared to brain
T2 compared to brain
Enhancement
Calcification
Acoustic neuroma Meningioma Epidermoid Metastasis
Hypo/iso Hypo/iso Iso to CSF Iso
Hyper Iso Iso/hyper to CSF Iso
Intense Intense Rare Moderate
Very rare Common Rare None
Source: From Ref. 3.
intracranial metastasis is present in 75% of cases (10). Additionally, patients with CPA metastases often have multiple cranial nerve involvement and meningeal disease (10). Table 2 compares MRI findings of common CPA masses.
PREOPERATIVE PREPARATION Careful preoperative preparation and evaluation of a patient with a CPA tumor incorporates clinical status, results from clinical investigations such as PTAs and speech discrimination tests, and diagnostic imaging study results. Much has been written about specific treatment paradigms of patients with acoustic neuromas (27–29). A substantial body of literature also exists regarding management of patients with CPA meningiomas (14,30,31). A thorough review of both of these topics is beyond the scope of this chapter.
We now present our general management scheme for patients with acoustic neuromas at our center (Fig. 7). Younger patients (≤40 years) are generally offered surgery with the goal of total excision. For this population, a retrosigmoid approach is selected for patients with small tumors (≤2.5 cm) and good hearing (≤50 dB on PTA and >50% speech discrimination) as well as for patients with large tumors (>2.5 cm) on the side of their only hearing ear. Younger patients with small tumors and poor hearing (>50 dB on PTA and ≤50% speech discrimination) as well as patients with large tumors will generally have a translabyrinthine approach for tumor resection. In middle-aged patients (41– 70 years) in otherwise good health, surgery is usually offered. The considerations in selecting the surgical approach are the same as the ones used for the younger patient population. However, although total surgical resection is still the goal, it is pursued less aggressively in this group than it is
Patient age
< 40 yr
< 41–70 yr
Surgery: Total excision Small tumor
Surgery: Total excision or subtotal resection + SRS
Large tumor
Retrosig
Small tumor
Large tumor
Serial MRI Small tumor
Good Poor In the only All hearing hearing hearing ear others Retrosig Translab
< 70 yr
Good Poor hearing hearing
Large tumor In the only All hearing ear others
Translab Retrosig Translab
Retrosig
No growth Follow-up MRI
Growth
Tumor debulking + SRS
SRS
Translab
LEGEND Small tumor: < 2.5 cm Large tumor: > 2.5 cm Poor hearing: > 50 dB on pure tone audiogram, < 50% speech discrimination SRS: Sterotactic radiosurgery Retrosig: Retrosigmoid craniectomy Translab: Translabyrinthine craniectomy MRI: Magnectic resonanace imaging
Figure 7
Treatment algorithm for patients with acoustic neuromas.
394
Oh et al.
with younger patients. If total surgical excision cannot be achieved, patients are then offered stereotactic radiosurgery. In older patients (>70 years), small tumors are followed with serial MRI scans. For patients with small acoustic neuromas that demonstrate growth, stereotactic radiosurgery is offered. Older patients with large tumors are treated with surgery for tumor debulking followed by stereotactic radiosurgery. The management paradigm for patients with CPA meningiomas is similar. In general, patients with meningiomas in the CPA also undergo a preoperative preparation process similar to patients with acoustic neuromas. In several instances, large cranial base meningiomas that also happen to involve the CPA require staged surgical resection or debulking followed by stereotactic radiosurgery. Finally, CPA epidermoids are usually treated with a retrosigmoid approach for surgical excision. There has been no demonstrated role for radiosurgery or radiotherapy in the treatment of epidermoids. In many instances, epidermoids are difficult to remove because of their adherence to adjacent nerves and vessels. The use of constant electromyographic facial nerve monitoring is now accepted as standard practice for surgery of acoustic neuromas and other CPA tumors. Electrodes are placed in the ipsilateral orbicularis oculi and orbicularis oris for the detection of muscle action potentials in response to surgical manipulation or monopolar or bipolar electrical stimulation of the facial nerve (2). For maximal benefit from continuous facial nerve electromyographic monitoring, it is preferable that a nonmuscle relaxant anesthetic technique is used. Perioperatively, a Foley catheter and arterial line are inserted. Antibiotics are given prior to skin incision. Dexamethasone, furosemide, and mannitol are also given intravenously prior to exposure and opening of the dura.
sen for right-sided tumors to prevent the patient’s shoulder from getting in the way of the surgeon’s right hand. The left retrosigmoid approach is performed with the patient in the supine position and head turned towards the right. In this case, the patient does not need to be in the lateral position because patient’s shoulder will not be in the way of the operating surgeon’s right hand. For an approach from either side, the patient’s head is held in place with a Mayfield threepoint fixation device. The abdomen is prepared for possible harvest of a fat graft. However, an abdominal fat graft is generally not needed for closure unless the patient has had a previous surgical approach to the CPA.
Incision and Soft Tissue Dissection A curvilinear incision centered 1 to 2 cm medial to the mastoid process is made. Scalp flaps are then developed with monopolar electrocautery and then elevated with fishhooks and rubber bands. The suboccipital fascia and muscles are then incised in a hockey-stick fashion and carefully separated from their attachments to the bone using subperiosteal dissection and monopolar electrocautery (Fig. 8). The fascia and muscles are also reflected with fishhooks. This two-layer opening facilitates better closure and may have a role in decreasing incidence of postoperative CSF leaks. An emissary vein is usually exposed in the region just medial to the mastoid process. Bleeding from this vein is stopped with bone wax.
SURGICAL TECHNIQUE Tumors of the CPA are generally approached by either a suboccipital retrosigmoid or translabyrinthine approach. Since an acoustic neuroma is by far the most common tumor of the CPA, further discussion of surgical technique will focus on this tumor. The microsurgical removal of an acoustic neuroma can be achieved via a suboccipital retrosigmoid, translabyrinthine, or middle fossa approach. Both the suboccipital and translabyrinthine approaches have been used to remove acoustic neuromas of all sizes (32). The middle fossa approach is used by some for small tumors in the IAC when an attempt is being made to preserve hearing (32). When hearing preservation is an issue, the suboccipital retrosigmoid approach is generally preferred. General considerations for selecting an operative approach have already been reviewed in the “Preoperative Preparation” section of this chapter. The retrosigmoid and translabyrinthine approaches for microsurgical excision of acoustic neuromas are now presented.
Retrosigmoid Approach Positioning The semi-sitting, prone, supine-oblique, lateral decubitus, and lateral oblique positions have all been used for suboccipital retrosigmoid removal of acoustic neuromas (32,33). At our center, the right retrosigmoid approach is performed with the patient in a left lateral decubitus position. The legs, hips, and arms are carefully padded. This approach is cho-
Figure 8 Illustration of the hockey-stick incision and the two-layer soft tissue dissection for a retrosigmoid or translabyrinthine craniectomy.
395
Chapter 26: Tumors of the Cerebellopontine Angle
Trigeminal nerve
Vestibulocochlear nerve
Tumor Facial nerve
Superior vestibular nerve
Tumor
Cerebellum
Figure 9 Extended retrosigmoid approach with dura opened. By drilling the bony constraints overlying the sigmoid sinus, the sigmoid sinus can be mobilized and the dura reflected more anteriorly allowing better exposure of more ventrolaterally located structures.
Inferior vestibular nerve Facial nerve
Bony Dissection The bony dissection for this approach involves a standard suboccipital craniectomy. A burr hole is placed behind the mastoid. Using a combination of Leksell and Kerrison rongeurs, the burr hole is enlarged. In areas where the bone is too thick for the rongeurs, a high-speed air drill with a diamond burr is used to thin the bone. The craniectomy is taken as high as the lateral sinus and as anterior as the sigmoid sinus. For an extended retrosigmoid approach, the bone is carefully drilled off the sigmoid sinus, exposing only a small amount of mastoid air cells. By decompressing the bony constraints of the sinus in this manner, better exposure is obtained ventrolaterally (Fig. 9). Exposed mastoid air cells, such as in an extended retrosigmoid approach or a partial mastoidectomy, should be thoroughly sealed off with bone wax. At this point, the operating microscope is brought into the field.
Tumor Resection If the dura remains tense at this point, additional measures are taken to relieve the tension. This includes lowering the patient’s pCO2 , temporarily raising the head of the operating table, or administering medications to induce additional diuresis. Once the dura is adequately relaxed, it is opened under high magnification in a curvilinear fashion and reflected anteriorly. A retractor is then gently placed on the cerebellum until the cisterna magna is identified. The cisterna magna is then sharply opened, which leads to the copious escape of CSF and, invariably, a recession of the cerebellum away from the posterior surface of the petrous bone (34). At this point, the tumor should be easily visible. The operative field should now be inspected while the tumor is still covered with its arachnoid investments, as inspection may reveal a branch of the AICA or AICA itself lying in a looplike manner across the cisternal portion of tumor (Fig. 10). By sharply cutting the arachnoid investments, it is possible to deliver the AICA out of the way of tumor dissection. In most cases, the vestibulocochlear nerve should now be easily identified splaying out into the medial aspect of the tumor while the facial nerve courses over the rostral aspect of the tumor. There is a difference of opinion as to whether an acoustic neuroma should first be decompressed internally or
Branch of anterior inferior cerenbellar artery
Figure 10 A branch of the AICA or AICA itself lie in a loop-like manner across the cisternal portion of an acoustic neuroma, as demonstrated by this intraoperative photograph. It needs to be delivered out of the field for surgical resection to proceed safely.
whether the IAC should be opened first so as to identify the cochlear and facial nerves in the canal. Our position as well as the position of Ciric et al (34) is to decompress a larger tumor first to deflect the medial edge of the tumor away from the bifurcation of the eighth cranial nerve into its cochlear and vestibular components. The decompression of the tumor is accomplished with pituitary forceps. The cochlear nerve can be significantly attenuated in patients with large tumors and may be barely recognizable even under high magnification. Once it is clear that the tumor is originating from the superior vestibular nerve, this nerve can be divided. As the tumor is decompressed, the identity of the cochlear nerve becomes progressively clearer to a point at where a plane may be visualized between the cochlear nerve and the tumor capsule (Fig. 11). Planes can then be developed between the tumor and the facial and inferior vestibular nerves. In all cases, it is important to avoid stretching or putting tension on the cochlear and facial nerves to prevent avulsion of the fibers. Once the cochlear nerve is identified, it is followed towards the porus acousticus. After the dura overlying the porus acousticus is stripped off, a 3-mm diamond burr is used to drill off the porus acousticus for approximately 3 to 4 mm to expose the dura of the IAC and thus the lateral extent of tumor. Excessive drilling may result in inadvertent opening of the posterior semicircular canal or endolymphatic duct and compromise hearing. The cochlear and facial nerves are once
396
Oh et al.
Translabyrinthine Approach Positioning
Inferior vestibular Superior vestibular Facial nerve
Figure 11 As an acoustic neuroma is internally decompressed, the identity of the cochlear nerve becomes progressively clearer to a point at where a plane may be visualized between the cochlear nerve and the tumor capsule as demonstrated by this intraoperative photograph.
The patient is placed supine on the operating table. After successful induction of general anesthesia, the patient’s head is placed on a donut turned so that the side of the tumor is facing up. The abdomen, ear, and postauricular area are prepped and draped in a sterile fashion. Patient position for a left translabyrinthine approach is shown in Figure 14.
Incision and Soft Tissue Dissection A curvilinear incision is marked out that starts from just below the tip of the mastoid process, runs upward over the lateral surface of the mastoid and just behind the root of the pinna, and courses to a point about 2 cm above the tip of the pinna (Fig. 15). Following incision, scalp flaps are developed using monopolar electrocautery and elevated with fishhooks and rubber bands. A hockey-stick incision is then made with the monopolar cautery in the temporalis fascia and muscle. The muscle and fascia are then separated off the
again identified within the IAC (Fig. 12). Under high magnification, tumor is then sharply dissected off of the nerves in the IAC and removed piecemeal. Bipolar forceps with lowcurrent settings may be used within the tumor capsule for coagulation. Outside the tumor capsule, it should be used very sparingly or not at all so as not to interrupt blood supply to the facial and cochlear nerves.
Closure Once hemostasis is achieved, the dura is closed over a piece of R suturable DuraGen (Integra LifeSciences, Plainsboro, NJ) in a watertight fashion. This closure is then reinforced with TISSEEL fibrin glue (Baxter Healthcare Corp., Deerfield, IL). Dural closure is demonstrated in Figure 13(A) to 13(D). A second piece of DuraGen is then placed on top of the construct and R once again reinforced with fibrin glue. A piece of Gelfoam (Pharmacia and Upjohn, Kalamazoo, MI) is then placed in the epidural space, and the cranial defect is repaired using a titanium mesh screen. The wound is then thoroughly irrigated with antibiotic irrigation. The fascia and muscle layer is then closed in an interrupted manner using absorbable suture. The scalp flaps are then reapproximated with interrupted absorbable suture. Finally, the skin is closed with a running, interlocking nylon suture.
(A)
(B)
Figure 12 Retrosigmoid approach after drilling of the internal acoustic canal, demonstrating the facial and vestibulocochlear nerve complex.
Figure 13 This series of pictures demonstrates dural closure of a suboccipital retrosigmoid approach for acoustic neuroma removal. (A) A piece of DuraGen is placed beneath the dural flap. (B) Dura on both sides of the durotomy is sutured to the underlying DuraGen. This maneuver helps to create a better seal. (C) A completed dural closure over the piece of DuraGen is demonstrated. (D) A layer of TISSEEL (fibrin glue) is placed over the construct.
Chapter 26: Tumors of the Cerebellopontine Angle
397
Figure 15 The incision to be made for a left translabyrinthine approach is marked out in this photograph. An incision is marked out that starts from just below the tip of the mastoid process, runs upward over the lateral surface of the mastoid and just behind the root of the pinna, and courses to a point about 2 cm above the tip of the pinna.
(C)
bone using subperiosteal dissection using monopolar electrocautery. The fascia and muscle are then elevated with fishhooks and rubber bands. A marking pen is then used to outline the approximate position of the sigmoid sinus (Fig. 16). This can be facilitated with the use of image guidance.
Bony Dissection A high-speed air drill with a cutting burr is then used to begin the bony dissection (Fig. 17). Anteriorly, the posterior wall of the external meatus is thinned. Superiorly, the dissection should expose the edge of the middle fossa dura and superior petrous sinus. The dissection is carried anteriorly above the external meatus as far as possible. The sigmoid (D)
Figure 13 (Continued.)
Asterion
Figure 14 Patient positioning for a left translabyrinthine approach is demonstrated.
Figure 16 In a translabyrinthine approach, the muscle and fascia are then separated off the bone using subperiosteal dissection using monopolar electrocautery. The fascia and muscle are then elevated with fishhooks and rubber bands. A marking pen is then used to outline the approximate position of the sigmoid sinus.
398
Oh et al.
Lateral canal
Jugular bulb Superior canal Posterior semicircular canal
Figure 17 A high-speed air drill with a cutting burr is then used to begin the bony dissection for a left translabyrinthine approach. (A)
sinus and roughly 1 cm of dura behind the sinus are exposed posteriorly. Inferiorly, the mastoid process is hollowed out. It is important to remove sufficient bone posteriorly and superiorly to improve access by retraction of the dura (2). With the aid of the operating microscope, the dissection is then deepened in the space between the middle fossa dura and superior margin of the meatus to open the mastoid antrum and atticus (Fig. 18). Following exposure of the incus and the head of the malleus, the incus is removed. The lateral semicircular canal is then identified on the medial wall of the epitympanic recess [Fig. 19(A) and 19(B)]. Dense bone marks the otic capsule surrounding the semicircular canals. The lateral semicircular canal is removed first. The superior canal is then drilled until the identification of the common crus, which can then be used to find the posterior canal. After completion of the labyrinthectomy, the bone covering the IAC is completely skeletonized over 270 degrees. The bone over the porus acousticus is then removed. The cavity resulting from this dissection is roughly pyramidal in shape, with its base at the cortical opening of the mastoid. The dura of the posterior fossa lies posteriorly. Dura of the middle fossa lies above, and the petrous bone, middle ear cavity, and descending facial nerve lie anteriorly.
Facial nerve
Canals opened
(B)
Figure 19 (A) Transmastoidal bone work prior to beginning of labyrinthectomy demonstrating the lateral, superior, and posterior semicircular canals. (B) Opening of the canals and the relationship of these structures to the facial nerve.
Tumor Resection
Sigmoid sinus
Antrum
Figure 18 Bony dissection of the sigmoid sinus and mastoid antrum are demonstrated. The sigmoid sinus is labeled in the photograph. The tip of the sucker is in the mastoid antrum.
To begin the next phase of the operation, the dura of the meatus and posterior fossa is opened. The extent of the incision in the posterior fossa is dependent on tumor size. Figure 20 shows a view of the surgical field immediately after the dura is opened. At this point, it is important to first identify the facial nerve in the internal auditory meatus. Once this is achieved, the tumor is then freed from its attachments to the arachnoid and dura at the porus, thus increasing the mobility of the entire mass. The arachnoid is then opened and a plane between the membrane and the surface of the tumor is established. Both the seventh and eighth nerves are then stimulated with a nerve stimulator to confirm their identities. The tumor is then gradually debulked. In nearly every case, it is not prudent to attempt tumor removal in one piece as the body of the tumor often obscures attachment to the facial nerve or internal auditory artery. By using the technique of internal tumor debulking, collapse of the tumor capsule is facilitated.
Chapter 26: Tumors of the Cerebellopontine Angle
Anterior/earlobe Acoustic neuroma Dura
Cranial
Caudal
399
communication. Eustachian tube obliteration is performed by drilling out the facial recess and subsequently packing muscle into the facial recess, out the tubotympanum into the Eustachian tube. Thin strips of harvested abdominal fat graft are placed into the dural defect in the region of the IAC under direct facial nerve monitoring. A larger piece of fat is placed on top of this construct. A piece of titanium mesh is then fashioned to cover the cranial defect. After the fascia and muscle, scalp flaps, and skin are then each closed in separate layers, a sterile dressing is applied.
POSTOPERATIVE CARE
Cerebellum
Posterior
Figure 20 This is the surgical field from a left translabyrinthine approach right after the opening of the dura. The picture is labeled for purposes of orientation.
The collapse allows the dissection to be extended under direct vision around the far side of the tumor, where arteries, cranial nerves, and the brainstem lie. As the tumor is reduced in size, dissection proceeds along its upper, lower, anterior, and posterior poles. Figure 21 displays a view of the surgical field after some tumor debulking and dissection have occurred. After tumor removal is completed, hemostasis is obtained. However, bipolar electrocautery is kept to a minimum during this process.
The patient is then extubated in the operating room and immediately taken to the intensive care unit. If the facial nerve is anatomically intact at the end of the operation, the facial musculature should be examined as soon as the patient recovers from anesthesia. If there are no complications, invasive monitoring devices are removed on the first postoperative day and the patient is transferred out to the ward. Oral feedings are also started on this day. Over the next few days, the patient is mobilized. Once adequately mobile and no complications are detected, the patient is sent home.
COMPLICATIONS AND COMPLICATION AVOIDANCE Postoperative Hematoma Hematoma in the postoperative cavity is exceedingly rare at our institution as well as at most other high-volume centers. If the patient does not recover promptly from anesthesia, or if there is an unexpected neurologic deficit or delayed mental status deterioration, a CT scan must be immediately performed to rule-out a cerebellar hematoma. Prompt removal of a hematoma can lead to a dramatic recovery and can be lifesaving.
Hydrocephalus
Closure Once hemostasis is achieved, the edges of the dural opening into the posterior fossa are approximated. The most important point of the closure is to seal off the middle ear from CSF
Hydrocephalus is another rare complication of surgery of the CPA. At our center, we have had one case of postoperative hydrocephalus after removal of an acoustic neuroma. If the patient develops signs and symptoms of hydrocephalus or if a tense subgaleal fluid collection begins to accumulate, a CT scan is warranted. In most cases, hydrocephalus resolves spontaneously. Should hydrocephalus persist, a ventriculoperitoneal shunt is placed.
Porous acousticus
CSF Leak Trigeminal nerve Jugular bulb Temporal line
VIII
Tumor
Sigmoid
Figure 21 This is the surgical field from a left translabyrinthine approach after some tumor has been debulked. The fifth and eighth cranial nerves are clearly visible. The porus acousticus has been drilled away.
CSF leak is a complication of CPA surgery, regardless of the approach. It has been reported with a frequency of between 2% and 25% (35–37). Depending on the case, CSF leak may present as rhinorrhea or leakage from the wound. Although it is most commonly evident within a few days of surgery, it may be a delayed complication and may occur after the patient is discharged from the hospital. Upon discharge from the hospital, patients should be given specific instructions about recognizing and reporting CSF leaks. In most cases, placement of a lumbar drain is the primary treatment for CSF leak. Should the problem persist after the lumbar drain is placed, the patient may need to be taken back to surgery for reexploration and graft repacking.
Meningitis When there is postoperative fever along with headache and/or nuchal rigidity, the possibility of either bacterial or aseptic meningitis should be explored. In cases of epidermoid
400
Oh et al.
tumor excision, aseptic meningitis is more common. A CT with contrast or an MRI scan is performed to look for an area that might be harboring an infection. A lumbar puncture is then performed and the CSF sent for appropriate studies. The patient is then started on broad-spectrum, intravenous antibiotics. Subsequent patient management is guided by results from the CSF studies.
Vertigo Postoperatively, patients may experience symptoms of vertigo, however, this complication is usually transient. In the case of chronic or permanent vestibular nerve dysfunction, patients generally eventually accommodate with central compensation (37).
Facial Nerve Dysfunction As mentioned above, facial musculature should be examined immediately in patients with anatomically intact facial nerves as soon as the patient recovers from anesthesia. Should facial paralysis be noted, eye care is of great importance, especially if facial analgesia is present and the corneal reflex is absent. The patient should have ointment and artificial tears applied to the affected eye on a regular basis. Additionally, the patient should wear an eye chamber while sleeping for further eye protection. During follow-up, the patient is specifically examined for corneal ulcerations. If there is suggestion of any problems, immediate referral to an ophthalmologist is warranted. A severely damaged cornea may lead to incapacitating pain and even visual loss. Sampath et al (38) have summarized the results of facial nerve function from several large series of acoustic neuroma patients. The House–Brackmann facial nerve grading system is the standard system used to record facial nerve function (18). When facial paralysis does not recover, the patient may need to undergo a hypoglossal–facial anastomosis. Additionally, tarsorrhaphy and/or gold weight placement in the upper eyelid may need to be performed. REFERENCES 1. Rhoton AL Jr. The cerebellopontine angle and posterior fossa cranial nerves by the retrosigmoid approach. Neurosurgery. 2000;47(3)(Suppl):S93–129. 2. Kaye AH, Briggs RJS. Acoustic neurinoma (vestibular schwannoma). In: Kaye AH, Laws ER Jr, eds. Brain Tumors. 2nd ed. London, England: Churchill Livingstone, 2001:619–669. 3. Osborn AG, Rauschning W. Brain tumors and tumorlike masses: Classification and differential diagnosis. In: Osborn AG, ed. Diagnostic Neuroradiology. 1st ed. St. Louis, MO: Mosby,1994:401– 528. 4. Matsuno H, Rhoton AL Jr, Peace D. Microsurgical anatomy of the posterior fossa cisterns. Neurosurgery. 1988;23(1):58–80. 5. House WF. Translabyrinthine approach. In: House WF, Luetje CM, eds. Acoustic Tumors: II-Management. 1st ed. Baltimore, MD: University Park Press, 1979:43–87. 6. Rhoton AL Jr. Microsurgical anatomy of the brainstem surface facing an acoustic neuroma. Surg Neurol. 1986;25(4):326–639. 7. Matsushima T, Rhoton AL Jr, de Oliveira E, et al. Microsurgical anatomy of the veins of the posterior fossa. J Neurosurg. 1983;59(1):63–105. 8. Nakamura M, Roser F, Dormiani M, et al. Facial and cochlear nerve function after surgery of cerebellopontine angle meningiomas. Neurosurgery. 2005;57(1):77–90; discussion 77–90. 9. Voss NF, Vrionis FD, Heilman CB, et al. Meningiomas of the cerebellopontine angle. Surg Neurol. 2000;53(5):439–446; discussion 46–47.
10. Yuh WT, Mayr-Yuh NA, Koci TM, et al. Metastatic lesions involving the cerebellopontine angle. Am J Neuroradiol. 1993;14(1):99– 106. 11. Deb P, Sharma MC, Gaikwad S, et al. Cerebellopontine angle paraganglioma—Report of a case and review of literature. J Neurooncol. 2005;74(1):65–69. 12. Sade B, Mohr G, Dufour JJ. Cerebellopontine angle lipoma presenting with hemifacial spasm: Case report and review of the literature. J Otolaryngol. 2005;34(4):270–273. 13. Harner SG, Laws ER Jr. Clinical findings in patients with acoustic neurinoma. Mayo Clin Proc. 1983;58(11):721–728. 14. Nakamura M, Roser F, Mirzai S, et al. Meningiomas of the internal auditory canal. Neurosurgery. 2004;55(1):119–127; discussion 27– 28. 15. Kobata H, Kondo A, Iwasaki K. Cerebellopontine angle epidermoids presenting with cranial nerve hyperactive dysfunction: Pathogenesis and long-term surgical results in 30 patients. Neurosurgery. 2002;50(2):276–285; discussion 85–86. 16. Meng L, Yuguang L, Feng L, et al. Cerebellopontine angle epidermoids presenting with trigeminal neuralgia. J Clin Neurosci. 2005;12(7):784–786. 17. Kavar B, Kaye AH. Dermoid, epidermoid, and neurenteric cysts. In: Kaye AH, Laws ER Jr, eds. Brain Tumors. 2nd ed. London, England: Churchill Livingstone, 2001:965–981. 18. House JW, Brackmann DE. Facial nerve grading system. Otolaryngol Head Neck Surg. 1985;93(2):146–147. 19. Johnson EW. Auditory test results in 500 cases of acoustic neuroma. Arch Otolaryngol. 1977;103(3):152–158. 20. Harner SG, Reese DF. Roentgenographic diagnosis of acoustic neurinoma. Laryngoscope. 1984;94(3):306–309. 21. Wu EH, Tang YS, Zhang YT, et al. CT in diagnosis of acoustic neuromas. Am J Neuroradiol. 1986;7(4):645–650. 22. Aoki S, Sasaki Y, Machida T, et al. Contrast-enhanced MR images in patients with meningioma: Importance of enhancement of the dura adjacent to the tumor. Am J Neuroradiol. 1990;11(5):935– 938. 23. Moller A, Hatam A, Olivecrona H. Diagnosis of acoustic neuroma with computed tomography. Neuroradiology. 1978;17(1):25– 30. 24. Thomsen J, Klinken L, Tos M. Calcified acoustic neurinoma. J Laryngol Otol. 1984;98(7):727–732. 25. Rosenberg SI. Natural history of acoustic neuromas. Laryngoscope. 2000;110(4):497–508. 26. Liu P, Saida Y, Yoshioka H, et al. MR imaging of epidermoids at the cerebellopontine angle. Magn Reson Med Sci. 2003;2(3):109– 115. 27. Rutherford SA, King AT. Vestibular schwannoma management: What is the ‘best’ option? Br J Neurosurg. 2005;19(4):309– 316. 28. Kondziolka D, Lunsford LD, Flickinger JC. Acoustic tumors: Operation versus radiation–making sense of opposing viewpoints. Part II. Acoustic neuromas: Sorting out management options. Clin Neurosurg. 2003;50:313–328. 29. Mangham CA Jr. Retrosigmoid versus middle fossa surgery for small vestibular schwannomas. Laryngoscope. 2004;114(8):1455– 1461. 30. Rhoton AL Jr. Meningiomas of the cerebellopontine angle and foramen magnum. Neurosurg Clin N Am. 1994;5(2):349– 377. 31. Nakamura M, Roser F, Dormiani M, et al. Surgical treatment of cerebellopontine angle meningiomas in elderly patients. Acta Neurochir (Wien). 2005;147(6):603–609; discussion 9–10. 32. Ojemann RG. Suboccipital transmeatal approaches to vestibular schwannomas. In: Schmidek HH, Sweet WH, eds. Operative Neurosurgical Techniques.Vol.1.3rd ed. Philadelphia,PA: W.B. Saunders Company, 1995:829–841. 33. Tew JM Jr, Scodary DJ. Neoplastic disoders-surgical positioning. In: Apuzzo MLJ, ed. Brain Surgery Complication Avoidance and Management. 1st ed. New York, NY: Churchill Livingstone, 1993:1609–1620. 34. Ciric I, Zhao JC, Rosenblatt S, et al. Suboccipital retrosigmoid approach for removal of vestibular schwannomas:
Chapter 26: Tumors of the Cerebellopontine Angle Facial nerve function and hearing preservation. Neurosurgery. 2005;56(3):560–570; discussion 70. 35. Gjuric M, Wigand ME, Wolf SR. Enlarged middle fossa vestibular schwannoma surgery: Experience with 735 cases. Otol Neurotol. 2001;22(2):223–230; discussion 30–31. 36. Gormley WB, Sekhar LN, Wright DC, et al. Acoustic neuromas: Results of current surgical management. Neurosurgery. 1997;41(1):50–58; discussion 58–60.
401
37. Ojemann RG. Acoustic neuroma (vestibular schwannoma). In: Youmans JR, ed. Neurological Surgery. Vol. 4, 4th ed. Philadelphia, PA: W.B. Saunders Company, 1996:2841–2861. 38. Sampath P, Rini D, Long DM. Microanatomical variations in the cerebellopontine angle associated with vestibular schwannomas (acoustic neuromas): A retrospective study of 1006 consecutive cases. J Neurosurg. 2000;92(1):70– 78.
27 Tumors of the Jugular Foramen Samer Ayoubi, Badih Adada, and Ossama Al-Mefty
preservation during tumor removal. By contrast, medially positioned tumors (some meningiomas, some schwannomas, and some glomus tumors) displace the cranial nerves onto the lateral tumor surface, where they interpose between the surgeon and the tumor—an unfavorable location (3). The carotid artery passes anteromedial to the internal jugular vein to reach the carotid canal. At the level of the skull base, this artery runs anterior to the vein and is separated from it by the carotid ridge. It ascends a short distance in the canal (the vertical segment), and then turns at a right angle anteromedially toward the petrous apex (the horizontal segment). Three branches of the external carotid artery—the ascending pharyngeal, the occipital artery, and the posterior auricular artery—can contribute significant blood supply to lesions of the jugular fossa. The sigmoid sinus (Fig. 3) courses down the sigmoid sulcus, turning anteriorly toward the jugular foramen, and crossing it into the jugular bulb. It then flows downward behind the carotid canal into the internal jugular vein. The inferior petrosal sinus courses on the surface of the petroclival fissure, forming a plexiform confluence as it enters the petrosal part of the jugular fossa. The position of the lower cranial nerves with respect to the inferior petrosal sinus varies; therefore, overpacking or cautery over the sinus can injure these nerves (6).
INTRODUCTION The jugular fossa is an area of complex anatomy. It is also an area of variant pathologies, each warranting special surgical considerations. Because lesions in this area involve the lower cranial nerves and major venous channels, each patient needs an individual approach that takes into account the location, size, and pathology of the lesion, as well as the patient’s general and neurological condition.
SURGICAL ANATOMY The jugular foramen is an opening or gap in the skull that connects the posterior cranial fossa and the jugular fossa (1). It is configured around the sigmoid and inferior petrosal sinuses between the temporal and occipital bones, and extends in a posterolateral-to-anteromedial direction. The jugular foramen hosts two venous compartments: the sigmoid part, which receives flow from the sigmoid sinus and the petrosal part, which receives drainage from the inferior petrosal sinus. A fibro-osseous diaphragm separates these two vascular channels, and the lower cranial nerves lie on either side of this partition at the site of the intrajugular processes of the temporal and occipital bones (2–4) [Fig. 1(A)]. The jugular fossa is a deep depression located at the inferior surface of the petrous portion of the temporal bone, and it communicates with the posterior cranial fossa via the jugular foramen. It houses the jugular bulb, which continues as the jugular vein inferiorly [Fig. 1(B)]. The 9th, 10th and 11th cranial nerves enter the dura on the medial side of the intrajugular process. The entrance porus of the glossopharyngeal nerve is separated from the entrance of the vagus and accessory nerves by a dural crest in the jugular fossa (5). The glossopharyngeal nerve (Fig. 2) passes forward, coursing through the jugular fossa, and exiting on the lateral surface of the internal carotid artery deep to the styloid process. The vagus nerve exits the fossa vertically, in intimate relation with the accessory nerve behind the glossopharyngeal nerve on the posteromedial wall of the internal jugular vein (4). The accessory nerve descends laterally between the carotid artery and the internal jugular vein and then backward across the lateral surface of the vein to the muscle. The hypoglossal nerve passes through the hypoglossal canal and does not traverse the jugular foramen. It passes adjacent to the vagus nerve and descends between the internal carotid artery and the jugular vein. Then, it turns abruptly forward toward the tongue. Lesions located lateral to the fibro-osseous diaphragm, which divides the two vascular channels, include glomus tumors, some meningiomas, and some schwannomas. These displace the nerves medially, a position favorable for nerve
REGIONAL PATHOLOGY AND DIFFERENTIAL DIAGNOSIS A variety of lesions can arise from the structures normally found within the jugular foramen and fossa or from contiguous structures. But the surgeon must be able to recognize anatomical variations of the jugular bulb, particularly a high jugular bulb or turbulent flow, so as not to misdiagnose them as a jugular foramen tumor (Fig. 4). Although rare, the three most common mass lesions within the jugular foramen are paragangliomas (including glomus jugulare tumors), schwannomas, and meningiomas (3,7–10) (Fig. 5). A firm preoperative diagnosis of these lesions is crucial because each has different surgical considerations, such as preoperative embolization for glomus jugulare tumors or the amount of bone to remove for meningiomas. Magnetic resonance imaging (MRI) complemented by computed tomography (CT) studies allow differentiation between these tumors (Table 1). Differential diagnoses to be considered are acoustic schwannomas and lesions such as chordomas and chondrosarcomas, malignant tumors (carcinomas), metastases, peripheral primitive neuroectodermal tumors, cholesteatomas, chondromas, lymphangiomas, choroid plexus papillomas, salivary gland tumors, lipomas, aneurysmal bone cysts, hemangiopericytomas, plasmacytomas and inflammatory granulomas, pseudomasses such as normal 403
404
Ayoubi et al.
Figure 2 Anatomical dissection demonstrating the relationship of the jugular bulb to the carotid artery, facial nerve, and lower cranial nerves.
Figure 1 Photographs of dry anatomical specimens from the right side. (A) Anterior perspective, in which the view extends from the outside to the inside, delineating the jugular foramen (white arrows) and jugular fossa (black arrows). (B) Posterior perspective, in which the view extends from the inside to the outside, delineating the jugular foramen (black arrows) and jugular fossa (white arrows). Note the difference between the two perspectives. Source: From Ref. 1.
vascular asymmetry, a high jugular bulb or jugular diverticulum, and aneurysms of the petrous carotid artery (7,9,10–15). A jugular foramen abscess has also been reported (16).
GLOMUS JUGULARE TUMORS Pathology Glomus jugulare tumors arise from glomus bodies located in the dome of the jugular bulb (17). These tumors grow along the path of least resistance and can gain access to the subarachnoid space by penetrating the dura of the posterior fossa, growing along cranial nerves, or, less commonly, penetrating the dura of the middle fossa (18). Glomus jugulare tumors are uncommon tumors of the head and neck, accounting for only 0.03% of all neoplasms and 0.6% of head and neck tumors. Nonetheless, they are the most common neoplasms of the middle ear and second to vestibular schwannomas as the most common tumor involving the temporal bone. These tumors appear in patients in their second decade (or earlier) to the ninth decade, although most tumors
Figure 3 The course of the sigmoid sinus inside the temporal bone and down to the neck.
Chapter 27: Tumors of the Jugular Foramen
405
Table 2 Classifications of Glomus Tumors Type
Physical Findings
Glasscock-Jackson Classification of Glomus Tumors I Small tumor involving jugular bulb, middle ear, and mastoid. II Tumor extending under internal auditory canal may have intracranial extension. III Tumor extending into petrous apex may have intracranial extension. IV Tumor extending beyond petrous apex into clivus or intratemporal fossa may have intracranial extension. Fisch Classification of Glomus Tumors A Tumors limited to the middle ear cleft. B Tumors limited to the tympanomastoid area with no infralabyrinthine compartment involvement. C Tumors involving the infralabyrinthine compartment of the temporal bone and extending into the petrous apex. Tumors with an intracranial extension less than 2 cm in diameter. D1 Tumors with an intracranial extension greater than 2 cm in D2 diameter. Figure 4 MRI of a high jugular bulb that may be mistaken for a glomus jugulare tumor. (A) Axial view. (B) Coronal view. (C) Magnetic resonance venography.
Figure 5 The most common pathological lesions in the jugular foramen. (A) Paraganglioma. (B) Schwannoma. (C) Meningioma.
manifest in the fourth decade of life. There is no clear racial predilection, but glomus tumors seem to be more common among Caucasians. There is a marked predominance among females; women are affected three to six times more commonly than men, with a peak incidence during the fifth decade of life. Multiple paragangliomas are reported in more than 10% of cases. Familial cases, most of which involve
Source: Adapted from Refs. 20, 21.
fathers and daughters, have a much higher rate of multicentricity, up to 55%. Evidence supports an autosomal dominant inheritance pattern consistent with genomic imprinting and an association with the haplotype at chromosome band 11q23. Most multicentric tumors are carotid body tumors. Only a few cases of bilateral glomus jugulare tumors associated with carotid body tumors have been reported (19). With rare exceptions, tumors of the glomus jugulare are benign, slow-growing paragangliomas. The two established classifications of these tumors, those of Fisch (20) and Glasscock and Jackson (21), are based mainly on tumor size, with special emphasis on intracranial extension as a decisive factor for resectability (Table 2). A subgroup of glomus jugulare tumors is rarely encountered but presents a formidable challenge for treatment. Al-Mefty and colleagues (22) used the following criteria to describe this group.
r r r r r
Giant size (Fig. 6) Multiple paragangliomas (bilateral or ipsilateral) (Fig. 7) Malignancy (Fig. 8) Catecholamine secretion Association with other lesions such as a dural arteriovenous malformation or an adrenal tumor, or previous treatment with an adverse outcome that makes surgical intervention a much higher risk, such as sacrifice of the carotid artery, radiation therapy, or postoperative deficits or adverse effects from embolization.
Table 1 Imaging Differences According to Type of Tumor Tumor
CT
MRI
Glomus jugulare
Erosion and destruction of the jugular spine and carotid crest (carotico-jugular spine, the bone separating the petrous carotid from the jugulare bulb). Moth-eaten pattern Enlarged fossa with smooth, distinct, sclerotic margin.
Salt-and-pepper appearance. Nonhomogeneous enhancement
Schwannoma
Meningioma
Frequently involves bone (including the jugular spine and particularly the jugular tubercle), producing hyperostosis and bone thickening without bone erosion.
Low to isointense on T1-weighted image, high signal on T2. May be cystic. Often has dumbbell shape. Moderate-to-marked enhancement. Tail sign. Extensive enhancement.
406
Ayoubi et al.
Figure 9 Deviation of the soft palate with atrophy and deviation of the tongue as presenting symptoms in a patient with a jugular foramen lesion.
Figure 6 MRI.
Examples of a giant glomus jugulare tumor. (A) CT scan. (B)
Clinical Presentation Symptoms of glomus tumors usually include pulsatile tinnitus, hearing loss, and lower cranial nerve palsies (18). Conductive hearing loss is the result of mechanical obstruction of the ossicular mechanism by the tumor, while sensorineural hearing loss is a result of the involvement of the labyrinth. Lower cranial nerve deficits are the prominent feature in patients with symptomatic jugular tumors (22) (Fig. 9). Symptoms of catecholamine production include palpitations, excessive sweating, and headache. These symptoms should be considered in any patient with a glomus jugulare tumor. Large tumors can cause obstructive hydrocephalus with symptoms and signs of increased intracranial pressure.
Appearance on Diagnostic Imaging
Figure 7 Examples of multiple paragangliomas. (A) Angiogram. (B) MRI. Abbreviations: RCCA, right common carotid artery; LCCA, left common carotid artery.
Figure 8 Examples of a malignant paraganglioma. (A, B) MRI after previous treatment with surgery and radiotherapy. (C) MRI showing marked growth a year later. (D) MRI after resection. (E) MRI showing rapid growth after 4 months. (F) Immunochemical staining for chromogranin. Note the brown granules in the cytoplasm. Source: From Ref. 22.
Bone-window CT scans show skull-base infiltration with erosion and enlargement of the jugular foramen (23,24). The extensive bone destruction is characterized by an irregular “moth-eaten” pattern of erosion (Fig. 10). MRIs show the enhanced tumor with flow voids and a “salt-and-pepper” appearance (Fig. 11), and also disclose the presence of multiple tumors. MRI of the neck is done to exclude associated paragangliomas. The arteriographic findings of glomus jugulare tumors are typically a hypervascular mass with an intense, characteristic tumor “blush.” Large feeding vessels and early draining veins are commonly encountered, indicative of early arteriovenous shunting. Similar to other tumors in this region,
Figure 10 Bone destruction by a glomus jugulare tumor, as seen on the bone window of a CT scan.
Chapter 27: Tumors of the Jugular Foramen
407
Figure 13 Angiographic depiction of a new blood supply through the opposite internal carotid artery and the vertebrobasilar system in a patient who had prior treatment with embolization and carotid occlusion. (A) Opposite carotid injection. (B) Vertebral injection.
Figure 11 Contrast-enhanced MRI demonstrating the typical salt-andpepper enhancement.
speech evaluations and swallowing studies are conducted before surgery.
Hormonal Studies glomus jugulare tumors are predominantly supplied by the external carotid artery system, mainly the ascending pharyngeal artery (23) (Fig. 12). Angiographic studies are critical for assessing the appropriateness of preoperative embolization after the tumor’s blood supply has been delineated. The most critical aspect of the angiographic evaluation in patients who have undergone previous embolization or carotid occlusion is the identification of new feeding vessels from the internal carotid artery and the vertebrobasilar circulation (22) (Fig. 13).
Preoperative Preparation With some modifications, the same preoperative protocol is used for all lesions of the jugular foramen. Extensive audiological and otolaryngological evaluations are carried out, and
Paragangliomas have the potential to secrete a wide variety of neuropeptide hormones, including adrenocortical hormones, serotonin, catecholamines, and dopamine (22). Patients with hypersecreting tumors (catecholamine levels four times higher than normal) require preparation with combined alpha- and beta-blocker medication before surgery, angiography, or embolization. Beta-blockers should not be given before or without alpha-blockers. Screening for excess catecholamines is necessary in all patients with a glomus jugulare tumor, and the actual treatment and duration of prophylaxis depends on the level of catecholamine secretion and its source. Patients with these tumors might also harbor adrenal norepinephrine-secreting tumors. Therefore, adrenal imaging is part of our workup and is particularly important for patients with hypersecreting tumors.
Diagnostic Imaging MRI scans with and without contrast enhancement, angiography, or arteriovenous magnetic resonance angiography, and very thin slices of CT scans constitute the radiological workup needed to explore the anatomy of each patient’s jugular fossa, temporal bone, and condyles. The nature of the tumor (cystic or solid), extensions (intracranial, extracranial, or dumbbell), and the characteristics of bone involvement (the presence of sclerosis and enlargement of the canal) are studied with MRI images and CT scans with the aid of a bone algorithm. The dominant vertebral artery and the characteristics of the vertebrobasilar system are also studied. Special attention is paid to the venous phase to determine the size, dominance, and tributaries (superior petrosal, inferior petrosal, and vein of Labb´e) of the transverse and sigmoid sinuses, and the position and size of the jugular bulb. Figure 12 The typical appearance of a glomus jugulare tumor on arteriograms. (A) Internal carotid. (B) The external carotid image shows the tumor’s high vascularity, venous shunting, and blood supply through the ascending pharyngeal artery.
Embolization Embolization is indicated for patients with paragangliomas and other highly vascular lesions (7,25,26). Surgery is more challenging when the patient has undergone prior
408
Ayoubi et al.
Figure 14 Preoperative embolization is of great value in decreasing the tumor’s blood supply, particularly from the external carotid artery. (A) Preembolization angiogram. (B) Postembolization angiogram.
embolization or carotid artery occlusion, after which new feeding vessels from the internal carotid and vertebrobasilar circulation developed (22). Current techniques are successful for embolizing the tumor bed and reducing blood loss (Fig. 14). The thoroughness of embolization is critical; however, partial embolization of the external carotid feeder augments the internal carotid feeders. Furthermore, embolization has accompanying risks and complications, including reflux cerebral emboli in the internal carotid artery, cranial nerve deficits from a “dangerous anastomosis,” and tumor hemorrhage. Carrier and colleagues reported that preoperative embolization of the inferior petrosal sinus, the anterior condylar vein complex, and the posterior condylar vein reduced preoperative bleeding considerably (27). Because of the absence of shift at the rigid structures of the jugular foramen, intraoperative image-guided frameless navigation is particularly useful during the surgical procedure (Fig. 15).
Intraoperative Neurophysiological Monitoring The 10th cranial nerve is monitored intraoperatively with an electromyographic endotracheal tube. Electromyographic needles are inserted into the facial musculature, the sternocleidomastoid muscle, and the tongue to monitor the 7th, 11th, and 12th cranial nerves, respectively. Auditory evoked potentials are obtained if hearing is present and no plan is made for closing the ear canal or resecting tumor from the middle ear. The tube is placed under sterile conditions after the ear is prepared. Multiple paragangliomas present the greatest challenge to treating complex paragangliomas because the treatment decision is not based on a single tumor, but on the quality and length of the patient’s life. Whether to treat, when to treat, which tumors to treat, with which modality (surgery or radiation) and in which sequence are all questions that must be addressed at the first evaluation and thoroughly considered throughout the patient’s follow-up. The surgeon must try to prevent the consequences of multiple bilateral cranial nerve deficits (22).
Figure 15 Intraoperative navigation is a useful orientation tool for surgery in the foramen magnum. (A) MRI. (B) Intraoperative CT scan indicating the extension of the tumor to the top of the petrous apex.
data, the surgeon can then decide the most appropriate approach (28). The approaches include the infratemporal, combined infratemporal–posterior fossa, and combined approaches with a total petrosectomy (22).
Patient Position The patient is placed supine with the head elevated, turned away from the side of the lesion, and fixed in the three-point head frame. The abdomen and thigh are prepared for removal of fat and fascia lata grafts (Fig. 16).
Incision and Soft-Tissue Dissection An open C-shaped incision is made behind the ear and extended up to the temporal area and down transversely along the natural skin crease in the neck. In selected patients, in whom the middle ear is involved, the external ear canal is transected at the bony cartilaginous junction. The skin of the external ear canal is everted and closed as a blind sac. A small periosteal flap is kept attached to the skin flap and closed over the ear canal (Fig. 17). The skin flap, including the auricle, is
Surgical Approach The surgical approach is tailored in each patient according to the findings of preoperative imaging, the local anatomy, and the tumor’s characteristics and extension. Jugular foramen tumors with an intracranial extension should be carefully evaluated with regard to size, position, infiltrative potential, and vascularization. On the basis of this
Figure 16 Skin incision and patient positioning for glomus tumor surgery.
Chapter 27: Tumors of the Jugular Foramen
409
Figure 17 Skin dissection and closure of the ear canal. (Insert) Operative photograph. Source: From Ref. 49.
reflected anteriorly, and the sternomastoid muscle is detached from its insertion in the mastoid process. The neurovascular structures in the neck are dissected and exposed; these include the common carotid artery, internal carotid artery, external carotid artery, the jugular vein, and the cranial nerves (9th through 12th).
Figure 19 Skeletonizing of the facial nerve from the stylomastoid foramen to the geniculate ganglion.
Bone Removal The mastoidectomy is done with a high-speed drill, and bleeding is anticipated during drilling of the bone. The eardrum is removed and the tumor is located in the middle ear. The semicircular canals are exposed (Fig. 18), and the facial canal is located caudal to the lateral semicircular canal. The facial nerve is skeletonized from the stylomastoid foramen to the geniculate ganglion (Fig. 19). We have abandoned the practice of routine transposition of the facial nerve; after its skeletonization, we keep it in a tiny protective bony canal. If the nerve needs to be transposed, it is moved out of the fallopian canal and secured anteriorly. A radical mastoidectomy exposes the sigmoid sinus down to the jugular bulb and, in
Figure 18 The mastoidectomy and opening of the middle ear.
cases of intradural extension, is followed by a lateral and low posterior fossa craniectomy.
Tumor Isolation To isolate the tumor, the internal carotid artery and the jugular vein are followed upward toward the base of the skull. To expose the tumor, the posterior belly of the digastric muscle and the stylohyoid muscles is transected and the styloid process is removed. The ascending mandibular ramus is dislocated anteriorly, if necessary. The sigmoid sinus is then ligated distal to the tumor’s extension and proximal to the mastoid emissary vein. If the tumor extends into the middle ear or along the petrous carotid artery, the remnant of skin in the external ear canal is removed with the tympanic membrane. The internal carotid artery is exposed in the petrous canal through drilling of the bone over the carotid canal if the tumor has not already destroyed this bone. The eustachian tube is obliterated with a piece of muscle. The anterior pole of the tumor is then dissected from the internal carotid artery, and the small feeding arteries are coagulated with bipolar electrocautery. At this stage, the extradural tumor is completely exposed (Fig. 20). If the tumor does not extend into the middle ear, the approach should be modified to preserve both the middle and the inner ear. Thus, the tumor is exposed in the infralabyrinthine space. The superior pole of the tumor is freed from the infratemporal fossa. The inferior pole is removed by dissecting and elevating the jugular vein after it is ligated to prevent early venous drainage. The lower cranial nerves are preserved as they emerge from the jugular foramen. Intrabulbar dissection, a maneuver described by AlMefty and Teixeira (22), helps preserve the lower cranial nerves. This maneuver can be used for any tumor, as long as the tumor itself has not penetrated the wall of the jugular bulb or actually infiltrated the cranial nerves. The outer wall
410
Ayoubi et al.
Intradural Tumor Removal
Figure 20 Isolating the tumor. Source: From Ref. 22.
of the lower sigmoid sinus is incised along the jugular bulb into the jugular vein. The tumor is then removed from inside the jugular bulb and the sigmoid sinus, and the tail end of the jugular vein is separated from the lower cranial nerves. The innermost venous wall separating the tumor from the nerves is left in situ to minimize dissection, trauma, manipulation, or devascularization of the lower cranial nerves (Fig. 21). Using this technique helps preserve the immediate postoperative function of the lower cranial nerves for patients in whom the tumor does not transgress the venous wall at the jugular foramen. When the tumor does transgress the venous wall, the cranial nerves can be infiltrated on a microanatomical level despite having normal function (29). In these situations, total resection may not be possible without sacrificing these nerves (29). Profuse bleeding from the inferior petrosal artery is controlled through packing with appropriate hemostatic materials.
To remove the intradural portion of the tumor, the dura mater is excised posterior to the sigmoid sinus and carried forward, and the intradural extension of the tumor is exposed. The cranial nerves (8th through 12th) are meticulously dissected from the tumor and kept intact. Tumor encroachment on the medulla is removed through microdissection, and the basilar artery, the anteroinferior cerebellar artery (AICA), and the posteroinferior cerebellar artery (PICA) are dissected from the tumor and preserved. Any tumor extension into the foramen magnum is followed and removed after it is freed from the lateral and anterior surfaces of the medulla and the vertebrobasilar junction. When it is giant, the tumor should be isolated for safe surgical removal. This is best done in one stage through the combined posterior fossa and infratemporal approach described earlier (30). This approach allows the tumor to be devascularized from the intrapetrous carotid artery. It is also used to separate the tumor from the posterior fossa and to dissect the lower portion from the nerves with minimal blood loss while preserving the vessels. Resecting tumors of the glomus jugulare requires special techniques in the handling of both arterial and venous dissection. Although paragangliomas engulf, adhere to, and receive blood from the internal carotid artery, with the aid of the operating microscope, a plane of dissection can be identified to separate the tumor from the carotid. Thus, the carotid artery does not need to be sacrificed or reconstructed. Because exposing tumors of the glomus jugulare requires neck dissection, associated tumors of the carotid body can be removed at the same time without additional morbidity or undue lengthening of the operating time. Glomus jugulare tumors often shunt blood with high venous outflow. Accordingly, they should be handled as arteriovenous malformations. Therefore, venous drainage from the tumor should be preserved and the proximal end of the jugular vein should not be ligated until the tumor is isolated and its arterial supply is devascularized.
Closure Once total removal of the tumor is assured, the eustachian tube (if exposed) is covered with a small piece of muscle and fascia. The dura mater is repaired with a graft of fascia lata, and the cavity is obliterated with fat. The temporal muscle is swung inferiorly and sutured to the sternomastoid muscle, and the skin is closed in two layers (Fig. 22).
Postoperative Care The patient’s ability to swallow is tested postoperatively, and oral intake is allowed only if these studies show satisfactory results, which assure that the function of the lower cranial nerves can protect the airway. Otherwise, the patient is kept on the transpyloric feeding by Dobhoff tube until swallowing recovers. Adaptations are made for satisfactory airway protection and other appropriate precautions are taken. A CT scan is obtained during the early postoperative period to check for hemorrhage, hydrocephalus, edema, or infarction. Any residual tumor is better assessed on later contrast-enhanced and fat-suppression MRIs.
Reconstruction Figure 21 Intrabulbar dissection preserves the wall of the jugular bulb, if it is not involved by the tumor, and protects the lower cranial nerves.
Large surgical defects must be repaired with vascularized flaps (temporalis fascia, cervical fascia, sternocleidomastoid muscle, and temporalis muscle) (26). Reconstruction of the carotid artery as a graft bypass to the middle cerebral artery
Chapter 27: Tumors of the Jugular Foramen
411
Deficits of the lower cranial nerves are the main surgical complications. However, the success in maintaining function has alleviated many of these concerns, and vigilant percussion postoperatively has minimized pulmonary complications from aspiration until the patient adapts or vocal cord medialization is done. Thus, total resection is indicated and successful in treating complex glomus jugulare tumors despite the challenge encountered (22). The rarely encountered malignant type, however, carries a poor prognosis (22,33).
Other Treatments
Figure 22 Closing the dura and packing the space with a fat graft.
may be necessary if the carotid is injured beyond repair (26,31). Facial reconstruction with the greater auricular nerve, the sural nerve, or a 12/7 anastomosis might be needed if the facial nerve was transected (26).
Results The results of surgical treatment of glomus jugulare tumors have improved drastically with the advances in skull-base surgery. Surgical resection with long-term follow-up shows the effectiveness of total removal in achieving a cure (19,21,32) (Fig. 23). Even the most formidable tumors have been successfully resected with no mortality and low morbidity (22).
Radiation therapy has long been used to treat glomus tumors, particularly those that are only partially removed or have recurred. But glomus tumors are known to be radioresistant, and the effect of radiation is often the induction of fibrosis, mainly along the vessels supplying the tumor. Furthermore, persistent viable tumors are often present long after the patient undergoes radiation therapy. Radiation therapy has also been associated with long-term side effects that include osteonecrosis of the temporal bone, the development of a new malignancy, and demyelination. Early reports of stereotactic radiosurgery for glomus tumors are encouraging, and this modality may be useful in controlling symptoms. Radiosurgery appears to provide control if the target size is within the optimal size for treatment, and the preliminary results of this treatment suggest a symptomatic improvement of cranial nerve function. If it is proved effective in long-term control with few complications from cranial nerve deficits, radiosurgery will be a great complement to the current treatment of bilateral glomus jugulare tumors and residual lesions from the resection of giant tumors (17). In a review of papers published from 1994 to 2004, Gottfried and colleagues found that death and recurrence after treatment with either radiosurgery or surgery were infrequent, and concluded that both treatments could be considered safe and efficacious. Although associated with higher morbidity rates, surgery immediately and totally eliminates the tumor. The results of radiosurgery were promising, but the long-term recurrence rate is still unknown (34).
Complications In patients with glomus jugulare tumors, mortality is around 1%. Leaks of cerebrospinal fluid (CSF) occur in 8% of patients. Other complications include aspiration, wound infection, pneumonia, and meningitis (34). Postoperative CSF leakage is one of the most important complications in the surgery of jugular fossa tumors (35). A CSF leak can occur either from the skin or, more often, in the form of rhinorrhea. Meticulous closure is very important in preventing the leak. If a leak does occur, spinal drainage is carried out for 72 hours. If the leak continues, the wound is reexplored. If meningitis occurs as a result of a CSF leak, it is treated with antibiotics and spinal drainage.
Follow-up and Rehabilitation
Figure 23 Complete surgical resection is successful in the overwhelming number of patients with large or giant glomus tumors and provides the prospect of cure. (A) Preoperative CT scan of a giant glomus tumor. (B) Postoperative CT scan after total removal.
After surgery, each patient is followed up with MRI images at 3 months, 6 months, and then annually to detect any recurrence. Dysphonic symptoms are treated with medialization of the vocal cords (34,36). Most patients adapt to unilateral deficits of the lower cranial nerves; however, the older the patient, the longer the recovery. Hence, aspiration must be avoided and tube feeding is needed in the early stages, or a jejunostomy is done if aspiration persists for prolonged periods of time. Eye weights or tarsorrhaphy may be necessary if facial weakness is present (34).
412
Ayoubi et al.
SCHWANNOMAS Pathology Jugular foramen schwannomas are a rare pathological condition comprising 2% to 4% of intracranial schwannomas (7), with approximately 100 cases reported in the literature (37). Schwannomas located at the jugular foramen may arise from the glossopharyngeal, vagus, or accessory nerves (7,37,38). The proximity and clinical manifestation of schwannomas originating from the hypoglossal nerve are the reason some authors classify these tumors with jugular foramen schwannomas. These tumors can originate from the cisternal portion of the nerve and present with major intracranial growth, or from the foraminal portion expanding the bone, or from the distal portion and present with extracranial growth. Some appear with both extra- and intracranial growth through an enlarged jugular foramen (7). Pellet and colleagues (39) have classified jugular foramen schwannomas into four types.
r Type A: primarily intracranial, minimal extension into the jugular foramen
r Type B: primarily intraosseous, with or without an intracra-
Figure 24 Imaging studies of a jugular foramen schwannoma. (A) Preoperative axial MRI showing the dumbbell shape and enhancement. (B) Coronal T2-weighted image showing a hyperintense signal and a cystic component. (C) Preoperative sagittal MRI showing the extension into the upper trunk. (D) CT scan showing expansion and scalloping of the jugular foramen, which is typical of schwannomas. (E, F) Postoperative MRIs showing total removal.
nial extension
r Type C: primarily extracranial, minimal extension into the jugular foramen
r Type D: saddlebag- or dumbbell-shaped intra- and extracranial extensions
Clinical Presentation The primary symptoms of a jugular foramen schwannoma include dizziness, hearing loss, dysphagia, diplopia, tongue paresis, and hoarseness (38). Preoperative findings include mainly audiovestibular (hearing loss, tinnitus, and dizziness) and lower cranial nerve signs (dysphagia, hoarseness, weakness of the shoulder, and tongue paresis) (36–38). Almost all patients with dumbbell-shaped schwannomas have glossopharyngeal and vagal deficits. Hypoglossal and accessory nerve deficits appear in most patients, while hearing loss is less common (7). With large tumors, the abducens and facial nerves can be affected and patients may develop cerebellar signs or hydrocephalus. Careful evaluation of the lower cranial nerves is particularly important in jugular foramen schwannomas because they are the source of most life-threatening postoperative complications.
Appearance on Diagnostic Imaging An enlarged jugular foramen with well-delineated sclerotic margins is seen on thin-cut CT scans with bone algorithms (7,23,40). Schwannomas usually expand the foramen without eroding it (41), and have a low to isointense signal on T1-weighted images and a high signal on T2-weighted images. The tumor shows moderate-to-marked enhancement after the injection of gadolinium (7,23,38,40,42). Cystic degeneration can occur, and is well delineated on MRIs (38,43) (Fig. 24). On conventional or magnetic resonance angiography, schwannomas are avascular (7).
Preoperative Preparation Diagnostic imaging and intraoperative neurophysiological monitoring are done in the same way as for patients with glomus tumors. Hormonal studies and embolization are not needed.
Surgical Approach The real challenge for neurosurgical treatment of these tumors is to preserve the function of the lower cranial nerves
while achieving radical resection and decreasing the risk of recurrence. A repeated operation drastically increases the chance of injury to the lower cranial nerves. These challenges are found especially in patients with dumbbell-shaped tumors, in which the nerves are at risk during resection of the intracranial, intraforaminal, and extracranial sections. Schwannomas differ from glomus tumors and meningiomas located within the jugular foramen and fossa because they compress rather than invade the jugular bulb, and the nerves of origin are positioned anterior to this structure. The labyrinth is also preserved as hearing might occasionally improve after removal of the tumor. Because schwannomas within the jugular foramen tend to displace the jugular bulb posteriorly, the suprajugular approach allows the surgeon to remove the tumor without opening the wall of the bulb. Even if preoperative studies reveal an absence of flow into the jugular bulb, the transjugular approach is not used because the sinus usually recovers its patency after decompression. The suprajugular approach is essentially a presigmoid infralabyrinthine route (Fig. 25). It is used if the tumor extends anteriorly to the jugular bulb. A postauricular incision is made, and the internal carotid artery; external carotid artery; internal jugular vein; and 9th, 10th, 11th, and 12th cranial nerves are identified in the cervical region. The sternomastoid muscle is dissected, mobilized, and reflected inferiorly. A mastoidectomy is followed by complete skeletonization of the sigmoid sinus, jugular bulb, and jugular vein. The presigmoid, infralabyrinthine space is exposed and the dura mater is identified superior to the patent jugular bulb and inferior to the labyrinth. After the cerebellomedullary cistern is opened, releasing CSF, the tumor is exposed and debulked. The lower cranial nerves (9th through 12th), the PICA, the AICA, and the vertebral artery are dissected away from the tumor through the arachnoid plane, and the lesion is radically removed. The removal of a schwannoma is accomplished without sectioning the ear canal, entering the middle ear, or transposing the facial nerve (7). Some surgeons transpose the facial nerve for selective circumstances, such as cases in which scar tissue from a previous operation impedes control of the carotid artery and safe removal, the tumor has a large extension anteriorly to the petrous apex, or the middle ear
Chapter 27: Tumors of the Jugular Foramen
413
Figure 26 (A) MRI of a small intraforaminal schwannoma. (B) Magnetic resonance venography image showing a single and dominant drainage through the corresponding jugular bulb. Radiosurgery might be preferable because of the increased risks presented by the venous configuration. Figure 25 The suprajugular approach usually used for jugular foramen schwannomas. Abbreviations: JB, jugular bulb; JV, jugular vein; SS, sigmoid sinus; SPS, superior petrosal sinus. Source: From Ref. 1.
is extensively involved (37,38,44). Since schwannomas are avascular and largely soft and suckable, we have not found a need to transpose the facial nerve. Dumbbell-shaped jugular foramen schwannomas present a special challenge, with the risk of injury to the lower cranial nerves intracranially, intraforaminally, and extracranially (7,31). However, these tumors can be removed without creating additional neurological deficits. Furthermore, the patient can be expected to recover function in the affected cranial nerves (7).
Postoperative Course In these patients, the acute development of postoperative deficits before the development of compensatory mechanisms requires careful attention. Speech pathology and otolaryngological evaluations with pre- and postoperative swallowing studies are obtained. Oral intake is withheld and parenteral nutrition is administered. Swallowing exercises and soft mechanical diets with swallowing precautions are prescribed if the patient exhibits a risk of aspiration. Vocal cord medialization is done if there is persistent dysphasia or aspiration (7).
Results Complete excision is achieved in the majority of patients with a schwannoma (7,26,38,43). Preoperative palsy of the 5th, 6th, 7th, 9th, 10th, and 12th nerves can improve after the removal of a jugular foramen schwannoma (7,38). Hearing can also improve (7,43).
Other Treatments Gamma-knife radiosurgery can be offered to patients who have small tumors or intact lower cranial nerve function, and those who have declined surgery. It is also considered for patients who have residual or recurrent tumors after microsurgical resection (45). Experience with radiosurgery is limited due to the rarity of these tumors. We prefer to reserve radiosurgery treatment for the rare patient in whom the venous anatomy presents a considerable risk, such as when the patient has a single functioning ipsilateral sigmoid sinus and jugular bulb (Fig. 26).
MENINGIOMAS Pathology Primary jugular fossa meningiomas are one of the rarest subgroups of meningioma, with fewer than 40 cases reported in the literature (1,46). They constitute 9% of jugular fossa tumors (9,36). These meningiomas presumably arise from arachnoid cells lining the jugular bulb in the jugular fossa (1). Women are affected more than men (1). The transitional type of meningioma is most commonly found, closely followed by the meningotheliomatous type and the less common psammomatous meningioma (47). Their intimate relationship with the lower cranial nerves and jugular bulb, their involvement of the temporal bone, and their tendency to extend intracranially and extracranially have traditionally made their removal fraught with difficulty (1). Thus, the surgeon needs to tailor the approach to the local anatomy (the tumor–neurovascular relationship).
Clinical Presentation The symptoms of a jugular foramen meningioma are similar to those of a schwannoma, with signs of lower cranial nerve deficits and altered hearing (1).
Appearance on Diagnostic Imaging A meningioma of the jugular foramen permeates and shows sclerotic changes on bone algorithm CT studies (Fig. 27). These lesions most often demonstrate isointense or low signal on T1-weighted MRIs and intermediate signal intensity on T2-weighted images. They also show avid, homogeneous enhancement on contrast MRIs and a dural tail (Fig. 28). Meningiomas of the jugular foramen may demonstrate a relatively more aggressive appearance than similar intracranial lesions, but most often retain their cerebrospinal fluid/vascular cleft with the brain parenchyma intracranially. Angiography shows avid arterial blushing and prolonged contrast retention into the venous phase (38). MRAs have mostly replaced conventional angiography as a preoperative test.
Preoperative Preparation The steps to prepare the patient for surgery are similar to those for patients with glomus tumors and schwannomas. Particular emphasis is given to the position, patency, and size of the jugular bulb as seen on magnetic resonance venography performed during both arterial and venous phases.
414
Ayoubi et al.
Figure 29 The three approaches used to remove jugular foramen meningiomas. Source: From Ref. 1. Figure 27 CT of a jugular foramen meningioma showing bone invasion with hyperostotic features, which is characteristic of meningiomas.
Surgical Approach For patients with meningiomas, the involved dura mater and the bone of the jugular fossa should be resected to minimize the chance of tumor recurrence (1). The surgical approach is tailored to the local anatomy of the tumor and its relation to the neurovascular structures. Three different routes can be used (Fig. 29).
r The suprajugular approach, a presigmoid, infralabyrinthine route, is chosen if the jugular bulb is patent and the tumor extends primarily anteriorly. r The retrojugular approach, a transcondylar, transtubercle, retrosigmoid route, is chosen if the jugular bulb is patent and the tumor extends primarily behind. r The transjugular approach, an infratemporal route, is chosen for patients in whom the jugular bulb is totally occluded by the tumor. The position, incision, and soft-tissue dissection for each approach are similar to those used for schwannomas.
The Suprajugular Approach In the suprajugular approach, a total mastoidectomy is done with complete skeletonization of the sigmoid sinus,
jugular bulb, and jugular vein. The jugular fossa is accessed in the presigmoid infralabyrinthine space. The dura mater located superior to the patent jugular bulb and inferior to the labyrinth is opened. CSF is released from the cerebellomedullary cistern, and the tumor is dissected away from the lower cranial nerves (9th through 11th), the PICA, the AICA, and the vertebral artery, while the arachnoidal surgical dissection planes are preserved. The tumor is debulked with suction and bipolar coagulation or with an ultrasonic aspirator. The procedure is completed with microsurgical radical resection of the tumor.
The Retrojugular Approach In the retrojugular approach, the suboccipital bone is exposed and a small, inferior, lateral suboccipital craniotomy is performed, followed by a mastoidectomy and complete skeletonization of the sigmoid sinus, jugular bulb, and jugular vein. Drilling approximately one-third of the condyle usually suffices for the exposure, and postoperative stabilization is not necessary. Attention is then turned to the jugular tubercle, which is completely drilled away to facilitate opening of the jugular fossa, which lodges the jugular bulb. With the aid of an operating microscope, the dura mater is incised along the posterior border of the sigmoid sinus (Fig. 30). The tumor is carefully separated from the medulla oblongata, lower cranial nerves, vertebral artery, and PICA along arachnoidal planes, and is followed toward the jugular fossa. Careful, meticulous dissection of the tumor from the jugular bulb and the wall of the jugular vein is important. Ultrasonic aspiration or suction and bipolar coagulation are used to debulk the tumor.
The Transjugular Approach
Figure 28 Contrast-enhanced MRI of a jugular foramen meningioma showing intense homogeneous enhancement and the dural tail characteristic of a meningioma. (A) Preoperative image. (B) Postoperative image.
For patients in which the meningioma has invaded the sinus and occupied the jugular bulb, a transjugular approach similar to that for a glomus tumor is used (Fig. 31). The neurovascular structures in the neck (9th through 12th cranial nerves, jugular vein, and carotid artery) are dissected and exposed. A radical mastoidectomy exposes the sigmoid sinus down to the jugular bulb and is followed by a posterior fossa craniotomy. The jugular vein is followed superiorly to the jugular bulb. To enlarge the exposure, the posterior belly of the digastric muscle and the stylohyoid muscle are transected
Chapter 27: Tumors of the Jugular Foramen
415
Figure 30 The retrojugular approach, which includes skeletonization of the sigmoid sinus and drilling of the jugular tuberculum. Source: From Ref. 1.
Figure 32 Illustration of the transjugular removal of a jugular foramen meningioma that invades and occupies the jugular bulb.
and the styloid process is removed. The sigmoid sinus and jugular vein are ligated at a location proximal to the mastoid emissary vein and distal to the tumor obstruction. The inferior pole of the tumor is then dissected off the internal carotid artery and the jugular vein. The extradural tumor is thus completely exposed (Fig. 32). Bleeding from the inferior petrosal sinus may be profuse and is controlled by packing with Gelfoam. With the aid of the microscope, the dura mater is opened posterior to the sigmoid sinus and carried forward. The intradural tumor extension is then exposed. Meticulous intradural dissection of the tumor, performed while maintaining the arachnoidal dissection planes, helps preserve the function of the lower cranial nerves and the vertebral artery, the PICA, and the AICA at the anterolateral surface of the medulla oblongata.
of a good outcome, provided extensive evaluation and appropriate tailoring of the operative approach is done.
Results Radical tumor removal can be achieved in 83% to 100% of patients with jugular fossa meningiomas (1,36,46). The most common complications are transient deficits of the lower cranial nerves, which resolve or are compensated for in all patients within 1 month (1,36). Therefore, jugular fossa meningiomas can be radically resected with the expectation
Other Treatments As in the case of schwannomas, experience with radiosurgery is limited due to the rarity of these tumors. The results, however, are expected to be the same as those of radiosurgery for basal meningiomas, when it is used for residual or recurrent tumors or as the primary treatment. Recent literature is expanding the data on the efficacy, technique, control rates, risks, and complications of this treatment. The goal of radiosurgery, however, is “control” and long-term results are not yet available. The average follow-up for the usual reported radiosurgery series is too short to ascertain significant control of this slow-growing tumor. The risks and complications of radiosurgery are not negligible and include seizures, brain edema, neurological deficits, cranial nerve deficits, and the potential for radiation-induced tumors or a new malignant progression. Some authors have reported a pattern of aggressive recurrence after years of control with radiosurgery (48). For tumors of the jugular foramen, we prefer to reserve radiosurgery for the rare patient in whom the venous anatomy presents a considerable risk, such as a single functioning ipsilateral sigmoid sinus and jugular bulb or recurrent tumor, or for patients who are unsuitable for or decline surgery.
REFERENCES
Figure 31 MRI of a patient with a jugular foramen meningioma that invaded and occupied the jugular bulb. This lesion was targeted through the transjugular approach. (A) Preoperative image. (B) Postoperative image.
1. Arnautovi´c KI, Al-Mefty O. Primary meningiomas of the jugular fossa. J Neurosurg. 2002;97:12–20. 2. Katsuta T, Rhoton AL Jr, Matsushima T. The jugular foramen. Microsurgical anatomy and operative approaches. Neurosurgery. 1997;41:149–201. 3. Lustig LR, Jackler RK. The variable relationship between the lower cranial nerves and jugular foramen tumors: Implications for neural preservation. Am J Otol. 1996;17:658–668. 4. Rhoton AL Jr. Jugular foramen. Neurosurgery. 2000;47(Suppl 3):S267–S285. 5. Ozveren MF, Ture U. The microsurgical anatomy of the glossopharyngeal nerve with respect to the jugular foramen lesions. Neurosurg Focus. 2004;17(E3):12–16. 6. Inserra MM, Pfiser M, Jackler RK. Anatomy involved in the jugular foramen approach for jugulotympanic paraganglioma resection. Neurosurg Focus. 2004;17(E6):41–44.
416
Ayoubi et al.
7. Kadri PAS, Al-Mefty O. Surgical treatment of dumbbellshaped jugular foramen schwannomas. Neurosurg Focus. 2004;17(E9):56–62. 8. Lee YB, Kim SH, Kim HT, et al. Jugular foramen neurilemmoma mimicking an intra-axial brainstem tumor. A case report. J Korean Med Sci. 1996;11:282–284. 9. Ramina R, Maniglia JJ, Fernandes YB, et al. Tumors of the jugular foramen. Diagnosis and management. Neurosurgery. 2005;57(1 Suppl): 59–68; discussion 59–68. 10. Tekdemir I, Tuccar E, Aslan A, et al. Comprehensive microsurgical anatomy of the jugular foramen and review of terminology. J Clin Neurosci. 2001;8:351–356. 11. Chao CK, Sheen TS, Lien HC, et al. Metastatic carcinoma to the jugular foramen. Otolaryngol Head Neck Surg. 2000;122:922– 923. 12. Harvey SA, Wiet RJ, Kazan R. Chondrosarcoma of the jugular foramen. Am J Otol. 1994;15:257–263. 13. Iwasaki S, Ito K, Takai Y, et al. Chondroid chordoma at the jugular foramen causing retrolabyrinthine lesions in both the cochlear and vestibular branches of the eighth cranial nerve. Ann Otol Rhinol Laryngol. 2004;113:82–86. 14. Prasanna AV, Muzumdar DP, Goel A. Lipoma in the region of the jugular foramen. Neurol India. 2003;51(1):77–78. 15. Yamazaki T, Kuroki T, Katsume M, et al. Peripheral primitive neuroectodermal tumor of the jugular foramen. Case report. Neurosurgery. 2002;51(5):1286–1289; discussion 1289. 16. Mirza S, Dutt SN, Irving RM. Jugular foramen abscess. Otol Neurotol. 2001;22(6):973–974. 17. Teixeira A, Al-Mefty O, Husain MM. Paragangliomas of the skull base. In: Berger MS, Prados MD, eds. Textbook of NeuroOncology. Philadelphia, PA: Elsevier, 2004:366–370. 18. Jackson CG, Kaylie DM, Coppit G, et al. Glomus jugulare tumors with intracranial extension. Neurosurg Focus. 2004;17(E7):45–50. 19. Borba LAB, Al-Mefty O. Paragangliomas of the skull base. Neurosurg Quart. 1995;5(4):256–277. 20. Fisch U, Fagan P, Valavanis A. The infratemporal fossa approach for the lateral skull base. Otolaryngol Clin North Am. 1984;17:513–552. 21. Jackson CG. Basic surgical principles of neurotologic skull base surgery. Laryngoscope. 1993;103:29–44. 22. Al-Mefty O, Teixeira A. Complex tumors of the glomus jugulare. Criteria, treatment, and outcome. J Neurosurg. 2002;97:1356– 1366. 23. Lowenheim H, Koerbel A, Ebner FH, et al. Differentiating imaging findings in primary and secondary tumors of the jugular foramen. Neurosurg Rev. 2006;29:1–11. 24. Oghalai JS, Leung MK, Jackler RK, et al. Transjugular craniotomy for the management of jugular foramen tumors with intracranial extension. Otol Neurotol. 2004;52:570–579. 25. Kinney SE. Glomus jugulare tumors with intracranial extension. Am J Otol. 1979;1:67–71. 26. Ramina R, Maniglia JJ, Fernandes YB, et al. Jugular foramen tumors. Diagnosis and management. Neurosurg Focus. 2004;17(2):E5. 27. Carrier DA, Arriaga MA, Gorum MJ, et al. Preoperative embolization of anastomoses of the jugular bulb. An adjuvant in jugular foramen surgery. AJNR. 1997;18:1252–1256. 28. Fisch U. Intracranial extension of jugular foramen tumors [Letter]. Otol Neurotol. 2004:25(6):1041; author reply 1041–1042.
29. Sen C, Hague K, Kacchara R, et al. Jugular foramen. Microscopic anatomic features and implications for neural preservation with reference to glomus tumors involving the temporal bone. Neurosurgery. 2001;48:838–847. 30. Al-Mefty O, Fox JL, Rifai A, et al. A combined infratemporal and posterior fossa approach for the removal of giant glomus tumors and chondrosarcomas. Surg Neurol. 1987;28:423–431. 31. Kawahara N, Sasaki T, Nibu K, et al. Dumbbell type jugular foramen meningioma extending both into the posterior cranial fossa and into the parapharyngeal space. Report of 2 cases with vascular reconstruction. Acta Neurochir. 1998;140(4):323–330. 32. Michael LM II, Robertson JH. Glomus jugulare tumors. Historical overview of the management of this disease. Neurosurg Focus. 2004;17(E1):1–5. 33. Bojrab DI, Bhansali SA, Glasscock ME III. Metastatic glomus jugulare. Long-term follow-up. Otolaryngol Head Neck Surg. 1991;104:261–264. 34. Gottfried ON, Liu JK, Couldwell WT. Comparison of radiosurgery and conventional surgery for the treatment of glomus jugulare tumors. Neurosurg Focus. 2004;17(E4):22–30. 35. Ramina R, Maniglia JJ, Paschoal JR, et al. Reconstruction of the cranial base in surgery for jugular foramen tumors. Neurosurgery. 2005;56:337–343. 36. Ramina R, Neto MC, Fernandes YB, et al. Meningiomas of the jugular foramen. Neurosurg Rev. 2006;29:55–60. 37. C ¸ okkeser Y, Brackmann DE, Fayad JN. Conservative facial nerve management in jugular foramen schwannomas. Am J Otol. 2000;21:270–274. 38. Wilson MA, Hillman TA, Wiggins RH, et al. Jugular foramen schwannomas. Diagnosis, management, and outcomes. Laryngoscope. 2005;115:1486–1492. 39. Pellet W, Cannoni M, Pech A. The widened transcochlear approach to jugular foramen tumors. J Neurosurg. 1988;69:887–894. 40. Eldevik OP, Gabrielsen TO, Jacobsen EA. Imaging findings in schwannomas of the jugular foramen. AJNR. 2000;21:1139–1144. 41. Flint D, Fagan P, Sheehy J. An intracranial vagal schwannoma without jugular foramen erosion or vagal dysfunction. Otolaryngol Head Neck Surg. 2005;132:507–508. 42. Valvassori G, Palacios E. Schwannoma of the jugular foramen. Ear Nose Throat. 1998;77:732. 43. Carvalho GA, Tatagiba M, Samii M. Cystic schwannomas of the jugular foramen. Clinical and surgical remarks [see comment]. Neurosurgery. 2000;46(3):560–566. 44. Sanna M, Falcioni M. Conservative facial nerve management in jugular foramen schwannomas [Comment]. Am J Otol. 2000;21(6):892. 45. Muthukumar N, Kondziolka D, Lunsford LD, et al. Stereotactic radiosurgery for jugular foramen schwannomas. Surg Neurol. 1999;52(2):172–179. 46. Gilbert ME, Shelton C, McDonald A, et al. Meningioma of the jugular foramen. Glomus jugulare mimic and surgical challenge. Laryngoscope. 2004;114(1):25–32. 47. Maloney TB, Brackmann DE, Lo WW. Meningiomas of the jugular foramen. Otolaryngol Head Neck Surg. 1992;106:128–136. 48. Couldwell WT, Cole CD, Al-Mefty O. Patterns of skull base meningioma progression after failed radiosurgery. J Neurosurg. 2007;106:30–35. 49. Al-Mefty O. Atlas of Meningiomas. New York:Lippincott-Raven Press, 1997.
28 Tumors of the Craniovertebral Junction Douglas Fox, Scott Wait, Steve Chang, G. Vini Khurana, Curtis A. Dickman, Volker K. H. Sonntag, and Robert F. Spetzler
arch of the atlas with the laminae of the axis. The occiput is connected to the atlas anteriorly and posteriorly by the atlanto-occipital membrane. The lateral border of the posterior membrane passes posteriorly to the vertebral artery and first cervical nerve root and may be ossified in this area (2). The tectorial membrane, alar ligaments, and apical ligament of the dens also help attach the occiput to the axis. The CVJ contains the upper spinal cord, caudal brain stem, cerebellum, and lower cranial and upper spinal nerves. The spinal cord blends into the medulla of the brainstem where the ventral rootlets emerge to form the first cervical nerve. Dorsal nerves cannot always be identified, and the sensory component of the first cervical nerve is not always present. Thus, the medulla occupies the foramen magnum. The upper cervical region is notable for the dentate ligaments, which are fibrous attachments between the spinal cord and spinal dura that are present midway between the ventral and dorsal cervical nerve rootlets. The accessory nerve has a cervical component that forms from rootlets that emerge from the spinal cord anterior to the dorsal cervical rootlets. The cerebellum surrounds the foramen magnum but normally does not occupy space or pass through the opening. The cerebellar tonsils although superior to the lateral edge of the foramen can herniate through the foramen in numerous conditions. The lower four cranial nerves are also in the region of the foramen magnum. The glossopharyngeal, vagus, and accessory nerve exit the jugular foramen and can often be identified due to their separation as they pierce the dura entering the jugular fossa. The spinal component of the accessory nerve is the only nerve that passes through the foramen magnum, arising from rootlets of the cervical spine. The spinal and cranial portions of the accessory nerve combine as it exits the jugular foramen. The hypoglossal nerve exits through a canal bearing its name lateral to the tubercle of the foramen magnum. The hypoglossal nerve often passes posteriorly to the vertebral artery and may be intimately associated with the posteroinferior cerebellar artery. The vertebral arteries and their branches constitute the major vascular structures in the region of the CVJ. The posterior spinal arteries arise from the vertebral artery, often from its extradural portion, and enter through the dural perforation with the parent artery. The posteroinferior cerebellar artery typically arises intradurally from the vertebral artery but may do so extradurally as well. The spinal arteries arise from both vertebral arteries and combine to form the anterior spinal artery, which descends through the foramen magnum. The dura of the foramen magnum is supplied by the meningeal branches of the vertebral artery. These branches are located extradurally. They also arise from the internal carotid artery circulation via the ascending pharyngeal artery, which enters through the hypoglossal canal.
INTRODUCTION The craniovertebral junction (CVJ) refers to the region of the occipital bone, which constitutes the foramen magnum and the atlas and axis vertebrae. This region encompasses the brainstem, upper cervical spinal cord, cranial and spinal nerves, and the vertebral arteries and their branches. These vital structures are closed together within a bony enclosure and are thus susceptible to lesions involving the area. Given the numerous structures involved, clinical manifestations are diverse. Physicians must be aware of the many symptoms that can relate to the region of the foramen magnum. Lesions of the CVJ can involve multiple structures, which create special problems that demand intense planning and focus from surgeons.
SURGICAL ANATOMY The bony structures of the CVJ include the occipital bone of the skull and the atlas and axis vertebra of the spine. This bony structure protects the brain stem, upper spinal cord, cranial and spinal nerves, and vertebral arteries (Fig. 1) (1). The occipital bone creates the foramen magnum through which the medulla passes. The occipital bone consists of a clival and a squamous portion. The latter includes the occipital crest, which connects with the falx cerebelli. The lateral edge of the foramen is created by the occipital condyles, which articulate the connection between skull and atlas. On the foramen, small anterior tubercles are the site of attachment of the alar ligament of the dens. The hypoglossal canal, which transmits the hypoglossal nerve, lies just lateral to this tubercle in the foramen. The atlas is the first cervical vertebra and has no vertebral body or spinous process. The atlas consists of two lateral masses connected by an anterior and posterior arch. The appearance of the axis is more typical of the rest of the vertebrae, except for the dens, which projects upward to articulate with the atlas. Synovial joints are present between the articulating surfaces of the occiput, atlas, and axis. The dens also has a synovial joint that articulates with the atlas anteriorly and with the transverse ligament posteriorly. The atlas and axis are connected by the anterior and posterior longitudinal ligaments, the cruciform ligament, and the articular capsules. The portion of the posterior longitudinal ligament that covers the cruciate ligament and dens as it extends upward to the clivus is referred to as the tectorial membrane (Fig. 2). The cruciate ligament has vertical and transverse segments that form a cross behind the dens. The transverse ligament attaches to the tubercles of the atlas and articulates with the dens. The ligamentum flavum also connects the posterior 417
418
Fox et al.
Figure 1 Cadaveric view of the craniovertebral junction. Source: Image courtesy of Mauro Ferreira, MD.
The venous anatomy in this area is also critical. The basilar plexus and occipital and marginal sinuses encircle the foramen magnum. An epidural venous plexus is located laterally in the CVJ and provides a medial border for the vertebral artery.
REGIONAL PATHOLOGY AND DIFFERENTIAL DIAGNOSES Tumors can arise from many different structures that occupy the CVJ. The rate of growth and location of the tumor determine the clinical presentation. The source of the tumor, bone or soft tissue, determines the imaging modality that will best detail the lesion. Extramedullary tumors of the CVJ are difficult to diagnose, often manifesting with symptoms that are difficult to localize. Most of these tumors are meningiomas, which are three times as common as neuromas. Meningiomas are most common in females, and this predilection holds for lesions in
Figure 2 Bony and ligamentous considerations of the axis and atlas. With permission from Barrow Neurological Institute.
the foramen magnum. Other lesions—including dermoids, teratomas, neurenteric cysts, arachnoid cysts, fatty tumors, and lesions of infectious etiology such as tuberculomas—can rarely be found at the CVJ. Neurofibromas of the CVJ are also rare except in the presence of neurofibromatosis, which can be associated with multiple lesions. Meningiomas can have many radiographic findings: hyperostosis, bony erosion, enlarged vascular channels in the bone, meningeal thickening, calcification, and hemorrhage. CT scanning is necessary to define the bony involvement of the lesion. Enlarged foramen and evidence of bony erosion may help in the determination between meningioma and neurilemmoma. MRI with and without gadolinium will normally show diffuse and intense contrast enhancement in meningiomas, and this will often be less prevalent with neurilemomas. Arterial and venous structures can also be detailed with MRI and MRA/MRV. Arterial encasement and displacement are important preoperative details, which may prevent complications during surgical resection. Extramedullary tumors are usually symptomatic at a younger age and affect males and females in equal number. The slow-growing nature of most of these lesions means that symptoms usually occur late when the lesion has reached a considerable size. Symptoms can include headaches, neck pain, cranial neuropathies, swallowing problems, balance disturbances, hydrocephalus, breathing problems, hoarseness, and hyperreflexia. Symptoms are often progressive but can have a remitting course that can confuse the evaluation. Malignant tumors in the region of the foramen magnum also must be considered. Chondrosarcomas are malignant tumors that can arise primarily from the sphenopetroclival junction or from malignant transformation of enchondromas (3). Most chondrosarcomas are low-grade lesions, but a more aggressive subtype, including dedifferentiated and mesenchymal varieties, portend a poor prognosis. Their rate of recurrence is high even if they appear histologically benign. Thus, surgeons must make every effort to achieve gross total resection. Repeat resection is warranted for residual or recurrent tumor. Recent reports suggest that the recurrencefree survival rate at 10 years is 32% (3). Radiosurgery is effective in the treatment of these lesions. Local control after proton radiotherapy is 70% at 5 years (4). Radiotherapy may increase the interval between resection
Chapter 28: Tumors of the Craniovertebral Junction
and recurrence (5). Radiosurgery combined with fractionated radiation for the treatment of chondrosarcomas is associated with high complication rates (6). Chemotherapy has demonstrated no efficacy in the treatment of these lesions. Chordomas are thought to arise from remnants of the notochord. Clival chordomas were first described by Virchow and Luschka in 1856. These tumors can occur anywhere along the neuraxis. About 50% are located in the sacrococcygeal region, 37% in the clivus/cranium, and 13% in the vertebral bodies. Half of the latter are in the cervical spine. Chordomas account for 3% to 4% of primary malignant bone tumors. They are the fourth most common primary malignancy in bone after osteosarcomas, chondrosarcomas, and Ewing sarcomas. Chordomas have a small peak in incidence that occurs during the first and second decades. Their incidence increases slowly and again peaks in the fifth to seventh decades. The mean age at diagnosis of cranial chordomas is 38 years. Cranial lesions occur equally in males and females. Three subtypes of chordomas have been described. First, conventional chordomas are characterized by a lobular architecture, vacuolated (physaliphorous) cytoplasm, and mucoid matrix. Pleomorphism and mitoses are uncommon. Second, chondroid chordomas are characterized by the Heffelfinger morphologic criteria, which consist of the presence of chondroid differentiation in near zones of more conventional chordoma (7). In a morphologically heterogeneous tumor, diffuse cytokeratin staining is necessary to diagnose chondroid chordoma (8). It occurs almost exclusively in the cranial region and at one-third the rate of conventional chordomas. Third, dedifferentiated chordomas are similar to chordomas or chondroid chordomas but with the addition of a secondary high-grade sarcomatous component. The sarcomas resemble malignant fibrous histiocytoma, but elements of fibrosarcoma, osteosarcoma, and high-grade chondrosarcoma have also been described. They compose 1% to 8% of all chordomas and may occur spontaneously or after radiation to a conventional chordoma. Their prognosis is exceedingly poor. Most patients die of tumor-related complications within one year (9). CT is the best modality to evaluate bony erosion and invasion (10). Chordomas typically lyse surrounding bone. T1-weighted MRIs with contrast show low-to-moderate signal intensity while T2-weighted MRIs show heterogeneous hyperintensity (11). Local recurrence is the rule and metastasis may occur. As many as 20% of patients have clinically evident metastasis, and as many as 40% have autopsy-evident metastasis (12). The tumors recur a mean of 2 to 3 years after treatment but can recur as late as 10 years thereafter (13). Complication and death from chordomas are primarily related to local mass effect and recurrence. As with most central neuraxis lesions, symptoms depend on the size and location of the tumor at diagnosis. In one report, visual difficulties (blurring or cranial nerve III, IV, or VI palsy) and headache were the most common symptoms occurring in 62% and 18% of patients with cranial chordomas, respectively. Sixth nerve palsy was the most common sign with an incidence of 40% (14). With the exception of patients younger than 5, who do poorly, prognosis is thought to be better with younger patients than with older. Young children typically have malignant pathologicappearing tumors and tend to die quickly (15,16). When 40 years is used as the cutoff, 5- and 10-year survival rates are 75% and 63%, respectively, for the younger group and 30% and 11%, respectively, for the older group (17). After aggressive surgical resection, proton-beam radiation is
419
the preferred treatment for these lesions. Control or shrinkage rates as high as 80% have been reported (18). Given their slow growth, chordomas respond minimally if at all to chemotherapy. Both osseous and soft-tissue tumors can be benign or malignant. Metastatic tumors are more common in this area than primary osseous tumors (19,20). Osteogenic, or bone-producing tumors, include osteoid osteomas and osteoblastomas. Osteoid osteomas are found five times more frequently in males than females and manifest with localized back pain (21). Aspirin often relieves the pain. Pain can be severe and operative excision can be curative. As many as 10% of osteoid osteomas can localize to the spine (22), usually involving the posterior elements and manifesting with intense pain. Because they are less than 2 cm, these lesions do not cause neurologic compromise. Tumors arising from the cartilaginous growth plates are called osteochondromas or enchondromas. Osteochondromas rarely involve the spine and rarely need resection. However, they can become large enough to cause enough pain and disfigurement to require surgical attention. When the vertebral column is involved, almost half of the cases involve the cervical spine with a predilection for C1 and C2. Spinal enchondromas are extremely rare, and only the rare case that undergoes malignant degeneration needs treatment. This situation occurs with multiple enchondromatosis and Maffuci syndrome, defined by enchondromatosis associated with soft-tissue hemangiomas. Eosinophilic granulomas result when the growth of reticuloendothelial cells is uncontrolled. Eosinophilic granulomas can affect the region of the CVJ. When symptomatic, it can be treated with curettage and possibly with chemotherapy and radiation (23). Multifocal eosinophilic granulomas can be treated with chemotherapy. Spontaneous regression is possible, but the lesion can also be reactivated. Conservative management with immobilization should be tried for patients presenting with pain although biopsy and curettage of the lesion are often performed to confirm the diagnosis of eosinophilic granulomas (24). Plasmacytomas can occur in the CVJ and usually present with pain and cranial nerve deficits (25). Plasmacytomas are B-cell lymphocytic tumors and precursors to the systemic disease of multiple myeloma. These tumors must be treated aggressively with multimodality treatments that include surgical resection, chemotherapy, and radiotherapy. Depending on the extent of the tumor and the completeness of resection, surgical stabilization may be needed. The success rates of painful pathological fractures treated with vertebroplasty or kyphoplasty are high (26,27). Outcome is based on the later occurrence of multiple myeloma, which has a poor prognosis. It is most common in skull base and spinal plasmacytomas (28). Intramedullary tumors involving the foramen magnum are most commonly glial in origin, with an astrocytic predominance in the pediatric population and an ependymal predominance in the adult population. Other less common tumors include cavernous malformations, hemangioblastomas, and melanocytomas. Before surgery is planned, lesions with an infectious etiology that constitute the diagnosis of myelitis must be excluded in the initial evaluation. Many tumors originating from the jugular foramen may extend to the CVJ. Most of these are schwannomas, glomus tumors, or paragangliomas. For the purposes of this review, paragangliomas and glomus tumors are considered under the term glomus tumors. Glomus tumors arise from the paraganglia cells situated near the jugular foramen. Pulsatile
420
Fox et al.
tinnitus and hearing loss are the most common presenting symptoms. Their close association with lower cranial nerves, important vascular structures, and the middle ear makes their surgical management complex, and the resultant morbidity may cause significant functional limitations. Glomus tumors have an incidence of 1 case per 30,000 individuals (29). Glomus jugulare and tympanicum tumors are the second most common temporal bone tumor. Females are affected two to five times more often than males (30–33). Familial syndromes account for 20% of glomus tumors. In those patients, multiple lesions occur 35% to 78% of the time compared to only 10% of sporadic cases (30,33–37). Patients with glomus tumors may have hypertension related to endocrinologically active tumors. One to 3% of tumors secrete a symptomatic amount of catecholamines (36,38). Norepinephrine is the most frequently produced vasoactive amine, although vasoactive substances such as dopamine have been reported (34,38–41). Catecholamine breakdown products greater than four to five times normal may be clinically apparent, thus requiring alpha- and betaadrenergic blockade before surgery to avoid labile blood pressure (38). Radiation and open surgical excision are both excellent management options, depending on an individual’s specific clinical situation and anatomy. A meta-analysis revealed a total resection rate of 88.2% at first surgery, a surgical control rate of 92.1%, a recurrence rate of 3.1%, and a mortality rate of 1.3%. Among patients who received radiosurgery as their primary treatment, 36.5% of the tumors shrank and 61.3% stabilized. Symptoms improved in 39% of patients, and 2.1% had a recurrence. The morbidity rate directly attributable to treatment was 8.5% (42).
CLINICAL ASSESSMENT Variability in presentation is the norm for lesions of the CVJ. Pain is the most common complaint for lesions, and sensory disturbances are also common. Diffuse symptoms and signs include pain and dysesthesias, cranial neuropathies, balance problems, swallowing dysfunction, gait abnormalities, nystagmus, atrophy in skeletal musculature, hyperreflexia, discoordination, and sphincter disturbance. Progressive pain that occurs at night and that worsens with motion may signify a bony lesion. Local pain may be related to compression of a specific nerve root or to the pathologic collapse of surrounding vertebral structures. Extramedullary tumors compress and often involve surrounding structures, leading to cranial neuropathies, myelopathy, hyperreflexia, atrophy of limb muscles, hydrocephalus, and pain. Intrinsic lesions often are reported by the patient as a tightness in the affected area, and reports of pain and sensory problems are common. When lesions occur in the medulla, breathing problems, nausea, and vomiting are frequent findings. Lesions with extension to the jugular fossa usually cause unilateral lower cranial nerve dysfunction. Large tumors can involve cranial nerves VII, VIII, and XII. Pulsatile tinnitus may differentiate a glomus tumor from a jugular foramen schwannoma. Other symptoms of jugular fossa lesions can include bleeding from the ear, pain in the mastoid region, and cerebellar and brainstem signs from external compression. The diagnosis of many lesions is delayed because the evaluation is focused on other organ systems that appear involved. Thus, the varied presentation of lesions in the area of the CVJ must be considered during the evaluation.
DIAGNOSTIC STUDIES Imaging of the CVJ must be able to define the soft tissue, bone, central nervous tissue, and vascular structures as well as the relationships among structures. Osseous anatomy is best identified with CT with sagittal and coronal reconstructions. Lesions causing erosion, remodeling, fracture, or even ossification can be identified with CT. The relationships among the occiput, atlas, and axis can also be evaluated with sagittal and coronal reconstructions. Given the ease of CT scanning, diagnostic radiography has mostly been abandoned. Dynamic flexion and extension xrays, however, are still used routinely to determine instability. Discussion of the stability of the CVJ is beyond the scope of this chapter. Nonetheless, understanding the relationships among the respective components of the bony CVJ is essential when pursuing tumor resections that may cause immediate or eventual instability. MRI of the CVJ is important in both diagnosis and in determination of an operative goal and plan. Relationships between the offending lesion and the critical nervous and vascular structures are well delineated with MRI. Angiography is reserved for special cases where embolization may be a preferred route prior to further treatment or there is a need to better define both the arterial and venous anatomy. Computed tomography angiography (CTA) provides excellent three-dimensional reconstruction of vascular and bony anatomy and their relationships. Soft tissue detail is not as apparent on CTA compared to MRI, but CTA may be helpful in surgical planning.
SURGICAL TECHNIQUE Transoral Approach and Its Extensions The most appropriate indication for this procedure is an extradural mass lesion in the anterior skull base (Figs. 3 and 4). The transoral approach allows access to the lower clivus, atlas, and axis. MRI and CT with sagittal and coronal reconstructions are performed preoperatively. A vascular study is often obtained to determine the vertebral anatomy, and CTA with coronal and sagittal reconstructions is preferred. Broad-spectrum antibiotics are administered before surgery and continued for 24 hours after surgery. The patient is positioned supine in a three-pin fixation system with the head slightly extended (Fig. 5). Preoperative tracheostomy is an option in patients with severe bulbar or respiratory symptoms, but it usually is unnecessary with newer-generation retraction devices. Retractors elevate the palate and uvula, retract the tongue and endotracheal tube caudally, and allow lateral exposure of the soft tissue (Fig.6). A microscope is used for visualization, and the surgeon positions himself or herself at the head of the patient. Once the patient is positioned, intraoperative fluoroscopy is used to determine spinal alignment and to confirm the cranial and caudal extent of the exposure created with the retraction system. The C1 tubercle should be palpated to identify the midline. A midline incision in the median raphe of the pharyngeal wall sharply continues through the mucosa, muscle, and anterior longitudinal ligament. Subperiosteal dissection is then used to expose the area of interest from the clivus through C2. Self-retaining retractor blades are used to hold the soft tissue laterally to obtain direct exposure of the lower clivus, C1, and C2. Thereafter, the direction of surgery depends on the pathology and its location. The inferior aspect of the C1 arch can be partially resected to identify the lateral aspect of the
Chapter 28: Tumors of the Craniovertebral Junction
421
Figure 3 Axial trajectory of surgical approaches to the craniovertebral junction and skull base (superior view). With permission from Barrow Neurological Institute.
odontoid process. The apex of the dens can be identified between the clivus and the superior aspect of the anterior arch of C1. The base of the dens is drilled posteriorly to the cortex, and a small Kerrison rongeur is used to complete the transection from the body of C2. Once the ligamentous attachments
Figure 4 Sagittal view of the transoral exposure from the lower clivus to the dens. With permission from Barrow Neurological Institute.
Figure 5 Lateral view with transoral retractor in place. With permission from Barrow Neurological Institute.
Figure 6 Sagittal view depicting transoral retractor (inset: surgeon’s view of operative field). With permission from Barrow Neurological Institute.
422
Fox et al.
are cut at the apex and the dens is free, it can be pulled caudally and ventrally. The C1 anterior arch can also be removed to expose the entire dens, which can then be removed piecemeal with a high-speed drill. Its removal must start at the apex as it becomes difficult if the dens is disconnected from C2. The transverse ligament and tectorial membrane may need to be removed to resect an extradural mass lesion. The use of standard microsurgical dissection techniques avoids the complication of a cerebrospinal fluid (CSF) leak. Closure is a single step encompassing all layers with a running 2–0 Vicryl suture. If the procedure is likely to be intradural or if an intraoperative CSF leak develops, lumbar drainage should be performed. If the dura must be closed or repaired, a fascia graft and fibrin glue are used to strengthen the closure. Once surgery is complete, an enteral feeding tube is placed under direct visualization. After one week of enteral feeding, patients are advanced to an oral diet, starting with clear liquids. The endotracheal tube remains until tongue and pharyngeal swelling decreases. The postoperative stability of the spine also must be determined. The patient must be in a rigid orthosis. If required, stabilization and fusion are performed posteriorly. When possible, fusion is limited to C1 and C2. Anterior bone grafts are associated with significant risks of displacement and infection; thus, they are rarely used. Extensions of the transoral approach can increase rostral exposure. A transoral-extended maxillotomy allows access to lesions extending from the sella to C2.
Transfacial Approaches These approaches provide a downward angle of approach to the CVJ (Fig. 7). The approaches can be divided into transnasal, transmaxillary, and transpalatal and are done with
Figure 8 Sagittal view of the suboccipital exposure. With permission from Barrow Neurological Institute.
the assistance of an experienced craniofacial surgeon. The angle of approach depends on the area of disease and its relation to the clivus, C1, and C2. Variations of the transfacial approach are based on extensions of the supraorbital bar. Retaining the medial orbits and midline nasal structures on the supraorbital bar improves access to the inferior clivus. Including the lateral orbital walls allows lateral displacement of the orbits to increase the lateral extent of the exposure. A cribriform osteotomy can be used to preserve olfaction and to help prevent a CSF leak. Transmaxillary and transpalatal approaches require a Weber–Ferguson incision with LeFort II and LeFort I osteotomies, respectively. The widened exposure of the clivus provides access to lesions that extend rostrally behind the sella and inferiorly to the upper cervical vertebrae. Transnasal approaches provide access to lesions of the anterior cranial fossa, nasopharynx, and clivus. The transnasal approach is best for clival lesions with predominant anterior growth. Large clival lesions that extend in multiple directions including posteriorly and inferiorly need a transmaxillary approach. A transpalatal approach will allow for access to the entire clivus without violating the nasal sinuses and is preferred for smaller lesions (43).
Retrosigmoid and Far Lateral Approaches
Figure 7 Approaches to the skull base. (A) Transfacial, (B) transmaxillofacial, (C) transoral/transpalatal. With permission from Barrow Neurological Institute.
The suboccipital approach is used to access lesions involving the foramen magnum (Fig. 8). The posterior aspect of the occipital bone is removed to provide access to the posterior cervicomedullary junction. Based on the extent of bony removal, access can extend from the cerebellar hemispheres to the CVJ. The lateral suboccipital craniotomy, more often described as a retrosigmoid craniotomy, allows access to the CPA. When a retrosigmoid craniotomy is extended to the foramen magnum, the posterolateral CVJ region can be accessed. The retrosigmoid approach is performed with the patient in a park-bench position or supine with the head turned. Lumbar drainage is used routinely to relax the brain and to prevent postoperative CSF leaks. Intraoperative
Chapter 28: Tumors of the Craniovertebral Junction
Figure 9
423
Intraoperative far lateral exposure showing tumor at foramen magnum displacing the lower cranial nerves.
monitoring is used during all lateral suboccipital exposures. It includes the monitoring of somatosensory-evoked potentials, auditory-evoked responses, the facial nerve, and often cranial nerves IX through XII. The ninth and tenth cranial nerves are monitored via electrodes along the endotracheal tube; this procedure should be discussed with the patient before intubation. Motor-evoked potentials are monitored when intraparenchymal tumors are resected. When large, tumors in the anterolateral region of the CVJ displace and rotate the spinal cord (Fig. 9). Spinal nerve rootlets and the spinal accessory nerve can be draped over the posterior aspect of a tumor and often must be separated from the lesion. For exposure of lateral and anterior lesions at the CVJ, far lateral and extreme lateral approaches are preferred. The far lateral approach is usually sufficient for exposure of lesions at the CVJ. The far lateral exposure requires less bony removal and may lessen the risk of future instability of the spine when compared to the extreme lateral approach. The vertebral artery requires less manipulation in the case of the far lateral exposure and thus limits the risk of vascular injury. For a far lateral exposure (Fig. 10), the patient is placed in the park-bench position with the head rotated 45 degrees and flexed anteroposteriorly and laterally (Fig. 11). The arm is held in a sling under the headholder (Fig. 12). A lumbar drain is routinely placed unless the patients have an obstructive lesion. This aids in relaxation of the brain as well as potential reduction in postoperative complication of CSF fistula. A paramedian incision is used, and midline is identified at the distal end of the incision. The muscle layers are incised with cautery. Subperiosteal dissection proceeds until the foramen magnum and C1 and C2 lamina are exposed. Fish hooks are then used to retract the soft tissue from the field of view. A Leyla bar is often used to provide retraction and is moved once the tissue has relaxed. The vertebral artery seldom needs to be dissected free. Rather, its location should be respected and care taken to avoid injury. Bony exposure is the most critical step in obtaining adequate exposure. The bone is removed with a high-speed drill and can be completed as a craniotomy or craniectomy. The opening extends from the asterion and exposes the boundaries of the sigmoid and transverse sinuses. The craniotomy continues to the foramen magnum. Rongeurs or a high-speed drill is then used to increase the lateral exposure. The lateral two-thirds of the occipital condyle can be safely resected
because the hypoglossal foramen resides in the anterior third of the condyle. Bony exposure should be as complete as needed to provide a trajectory that requires minimal, if any, retraction. Venous bleeding can be controlled by packing with hemostatic agents. The dura is opened in a curvilinear fashion and reflected laterally to allow dissection of the vertebral artery, lower cranial nerves, upper cervical rootlets, and superior dentate ligament. Sharp section of the arachnoid layer and superior dentate ligament allows gentle retraction of the upper spinal cord and lower brainstem. Use of the far lateral approach minimizes retraction during resection of lesions in this area. It also affords proximal and distal control of the vertebral artery if needed for complex vascular lesions. Removal of a lesion depends on various factors including its size, vascularity, and firmness. Microsurgical techniques are combined with ultrasonic aspiration to resect most tumors.
Figure 10 Sagittal view of the (A) transpetrosal, (B) retrosigmoid, and (C) far lateral exposure of the craniocervical junction. With permission from Barrow Neurological Institute.
424
Fox et al.
for large tumors that are difficult to manage through standard approaches. In transpetrosal and transcochlear approaches, the defect must be reconstructed with fat to help prevent CSF leaks. Tumors of the jugular fossa require a combined approach with a skull base otolaryngologist to perform the mastoidectomy and petrosectomy. When added to a retrosigmoid craniotomy, these key components provide access to lesions.
COMPLICATIONS Tumors of the skull base are inherently difficult to resect because vital structures are often distorted by or involved with the lesions. The varied complications can include CSF leaks, lower cranial nerve palsies, vertebral and spinal artery injuries, brainstem injury, hydrocephalus, and infection. Postoperative medical complications may be high in this population because these patients typically have significant disabilities before they undergo definitive procedures. Instability in the spinal axis must also be addressed and can lead to neurologic and vascular compromise that affects outcomes. Depending on the extent of resection of the CVJ, stabilization techniques may be necessary.
CONCLUSION
Figure 11 Lateral view of the patient position for the far lateral craniotomy. With permission from Barrow Neurological Institute.
When possible, tumor capsules and arachnoid planes are preserved to prevent injury to surrounding structures. Surgical adjuncts that increase exposure for tumors extending above the CVJ include the combined middle fossa approach, petrosectomy, and transcochlear approach (Fig. 3). These approaches increase exposure and may be warranted
Tumors of the CVJ are complex lesions by their location and varied pathologies. Treatment paradigms range from biopsy with adjuvant therapies through to aggressive resection with adjuvant therapies. The lesions can require various approaches depending on the size and location of the lesion, which must be tailored on a case-by-case basis. We have provided a general overview of the lesions that are present and their respective treatments. Surgical approaches have been based on the workhorses of retrosigmoid, far lateral, suboccipital, and transoral, with the variations and extensions that are possible but less frequently used. Given the complex anatomy of the region, tumors of the CVJ are difficult to treat. This chapter details the presentation; diagnosis; treatment, including approaches for surgical access; and outcomes of the tumors that can involve the CVJ. Overall, the treatment of lesions involving the CVJ remains complicated and a challenge to all skull base surgeons.
Acknowledgment We thank Mauro Ferreira, MD, for contributing the dissection photograph for Figure 1. REFERENCES
Figure 12 Anterior view of the patient position for the far lateral craniotomy. With permission from Barrow Neurological Institute.
1. De Oliveira E, Rhoton AL Jr, Peace D. Microsurgical anatomy of the region of the foramen magnum. Surg Neurol. 1985;24:293– 352. 2. Rhoton AL Jr, De Oliveira E. Anatomical basis of surgical approaches to the foramen magnum. In: Dickman CA, Spetzler RF, Sonntag VKH, eds. Surgery of the Craniovertebral Junction. New York, NY: Thieme, 1998; Ch 2. 3. Tzortzidis F, Elahi F, Wright DC, et al. Patient outcome at longterm follow-up after aggressive microsurgical resection of cranial base chondrosarcomas. Neurosurgery. 2006;58:1090–1098. 4. Hug EB, Slater JD. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. Neurosurg Clin N Am. 2000;11:627–638. 5. Noel G, Feuvret L, Ferrand R, et al. Radiotherapeutic factors in the management of cervical-basal chordomas and chondrosarcomas. Neurosurgery. 2004;55:1252–1260.
Chapter 28: Tumors of the Craniovertebral Junction 6. Krishnan S, Foote RL, Brown PD, et al. Radiosurgery for cranial base chordomas and chondrosarcomas. Neurosurgery. 2005;56:777–784. 7. Heffelfinger MJ, Dahlin DC, Maccarty CS, et al. Chordomas and cartilaginous tumors at the skull base. Cancer. 1973;32:410– 420. 8. Radner H, Katenkamp D, Reifenberger G, et al. New developments in the pathology of skull base tumors. Virchows Arch. 2001;438:321–335. 9. Barnes EL, Kapadia SB, Nemzek WR, et al. Biology of selected skull base tumors. In: Janecka IP, Tiedmann K, eds. Skull Base Surgery: Anatomy, Biology, and Technology. Philadelphia, PA: Lippincott-Raven, 1997:263–292. 10. Larson TC III, Houser OW, Laws ER Jr. Imaging of cranial chordomas. Mayo Clin Proc. 1987;62:886–893. 11. Meyers SP, Hirsch WL Jr, Curtin HD, et al. Chordomas of the skull base: MR features. Am J Neuroradiol. 1992;13:1627– 1636. 12. Laws E, Thapar K. Parasellar lesions other than pituitary adenomas. In: Powell M, Lightman SL, eds. Management of Pituitary Tumors: A Handbook. New York, NY: Churchill-Livingstone, 1996:175–222. 13. Amendola BE, Amendola MA, Oliver E, et al. Chordoma: Role of radiation therapy. Radiology. 1986;158:839–843. 14. Colli BO, Al Mefty O. Chordomas of the skull base: Follow-up review and prognostic factors. Neurosurg Focus. 2001;10:E1. 15. Coffin CM, Swanson PE, Wick MR, et al. Chordoma in childhood and adolescence. A clinicopathologic analysis of 12 cases. Arch Pathol Lab Med. 1993;117:927–933. 16. Borba LA, Al Mefty O, Mrak RE, et al. Cranial chordomas in children and adolescents. J Neurosurg. 1996;84:584–591. 17. Forsyth PA, Cascino TL, Shaw EG, et al. Intracranial chordomas: A clinicopathological and prognostic study of 51 cases. J Neurosurg.1993;78:741–747. 18. Chang SD, Martin DP, Lee E, et al. Stereotactic radiosurgery and hypofractionated stereotactic radiotherapy for residual or recurrent cranial base and cervical chordomas. Neurosurg Focus. 2001;10:E5. 19. Black P. Spinal metastasis: Current status and recommended guidelines for management. Neurosurgery. 1979;5:726–746. 20. Boland PJ, Lane JM, Sundaresan N. Metastatic disease of the spine. Clin Orthop Relat Res. 1982;169:95–102. 21. Goldstein GS, Dawson EG, Batzdorf U. Cervical osteoid osteoma: A cause of chronic upper back pain. Clin Orthop Relat Res. 1977;129:177–180. 22. Raskas DS, Graziano GP, Herzenberg JE, et al. Osteoid osteoma and osteoblastoma of the spine. J Spinal Disord. 1992;5:204– 211. 23. Fernando UL, Cabezudo JM, Porras LF, et al. Solitary eosinophilic granuloma of the cervicothoracic junction causing neurological deficit. Br J Neurosurg. 2003;17:178–181. 24. Bertram C, Madert J, Eggers C. Eosinophilic granuloma of the cervical spine. Spine. 2002;27:1408–1413.
425
25. Menezes AH, Traynelis VC, Fenoy AJ, et al. Honored guest presentation: Surgery at the crossroads: Craniocervical neoplasms. Clin Neurosurg. 2005;52:218–228. 26. Fourney DR, Schomer DF, Nader R, et al. Percutaneous vertebroplasty and kyphoplasty for painful vertebral body fractures in cancer patients. J Neurosurg. 2003;98:21–30. 27. Hentschel SJ, Burton AW, Fourney DR, et al. Percutaneous vertebroplasty and kyphoplasty performed at a cancer center: Refuting proposed contraindications. J Neurosurg Spine. 2005;2:436– 440. 28. Schwartz TH, Rhiew R, Isaacson SR, et al. Association between intracranial plasmacytoma and multiple myeloma: Clinicopathological outcome study. Neurosurgery. 2001;49:1039–1044. 29. Mariman EC, Van Beersum SE, Cremers CW, et al. Analysis of a second family with hereditary non-chromaffin paragangliomas locates the underlying gene at the proximal region of chromosome 11q. Hum Genet. 1993;91:357–361. 30. Spector GJ, Gado M, Ciralsky R, et al. Neurologic implications of glomus tumors in the head and neck. Laryngoscope. 1975;85:1387–1395. 31. Alford BR, Guilford FR. A comprehensive study of tumors of the glomus jugulare. Laryngoscope. 1962;72:765–805. 32. Greer JA, Cody TR, Weiland LH. Neoplasms of the temporal bone. J Otolaryngol. 1976;5:391–398. 33. Spector GJ, Ciralsky R, Maisel RH, et al. Multiple glomus tumors in the head and neck. Laryngoscope. 1975;85:1066–1075. 34. Blumenfeld J, Cohen N, Anwar M, et al. Hypertension and a tumor of the glomus jugulare region. Evidence for epinephrine biosynthesis. Am J Hypertens. 1993;6:382–387. 35. Hodge KM, Byers RM, Peters LJ. Paragangliomas of the head and neck. Arch Otolaryngol Head Neck Surg. 1988;114:872–877. 36. Netterville JL, Jackson CG, Miller FR, et al. Vagal paraganglioma: A review of 46 patients treated during a 20-year period. Arch Otolaryngol Head Neck Surg. 1998;124:1133–1140. 37. Van Der Mey AG, Maaswinkel-Mooy PD, Cornelisse CJ, et al. Genomic imprinting in hereditary glomus tumours: Evidence for new genetic theory. Lancet. 1989;2:1291–1294. 38. Schwaber MK, Glasscock ME, Nissen AJ, et al. Diagnosis and management of catecholamine secreting glomus tumors. Laryngoscope. 1984;94:1008–1015. 39. Azzarelli B, Felten S, Muller J, et al. Dopamine in paragangliomas of the glomus jugulare. Laryngoscope. 1988;98:573–578. 40. Blumenfeld JD, Cohen N, Laragh JH, et al. Hypertension and catecholamine biosynthesis associated with a glomus jugulare tumor. N Engl J Med. 1992;327:894–895. 41. Troughton RW, Fry D, Allison RS, et al. Depression, palpitations, and unilateral pulsatile tinnitus due to a dopamine-secreting glomus jugulare tumor. Am J Med. 1998;104:310–311. 42. Gottfried ON, Liu JK, Couldwell WT. Comparison of radiosurgery and conventional surgery for the treatment of glomus jugulare tumors. Neurosurg Focus. 2004;17:E4. 43. Beals SP, Joganic EF, Hamilton MG, et al. Posterior skull base transfacial approaches. Clin Plast Surg. 1995;22:491–511.
Section 3 Tumor-Specific Considerations
29 Squamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses Patrick Sheahan, Snehal G. Patel, and Jatin P. Shah
INTRODUCTION
to rise with the length of exposure and in inverse proportion to the age at first exposure. The risk is thought to be multiplicative in nickel refinery workers who also smoke (15). Exposure to wood dust is well-established as a risk factor for adenocarcinoma of the ethmoid sinuses (16). The association between wood dust and SCC is less strong, however, Luce noted an eightfold increased risk of SCC in carpenters and joiners who had worked in the wood manufacturing industry for more than 15 years (17). Other authors have also reported a moderately increased risk for SCC among patients with long histories of wood dust exposure (18,19). There is some evidence that while exposure to hardwood dust is strongly associated with adenocarcinoma, exposure to softwood dust alone may be a risk factor for SCC (20). Other occupations associated with increased risk of nasal or paranasal sinus SCC include electrical workers (16), bakers and pastry cooks (17), grain millers (17), textile workers (21), farm workers (17), construction workers (17). Occupational exposure to asbestos (22) and formaldehyde (22,23) has also been implicated. Unlike SCC of the rest of the upper aerodigestive tract, tobacco and alcohol are traditionally not considered to be major risk factors for sinonasal cancer. However, there is good evidence from several studies that heavy smokers as well as snuff users do have an increased risk of SCC of the nose and paranasal sinuses (7,19,24,25). The risk would appear to be greatest in recent smokers (26). Radiation exposure would appear to be another factor capable of causing sinonasal SCC. Thorotrast (thorium dioxide) is a radiopaque material that was used as a contrast agent as recently as 1954. Upon injection into the maxillary sinus, it decays to mesothorium with concomitant release of alpha, beta, and gamma rays. Thorotrast appears to be retained in the maxillary sinus for life and radioactivity reaches a peak after 15 years. The link between thorotrast and sinonasal SCC is based on a series of case reports documenting maxillary sinus SCC in patients who had undergone thorotrast injection 10 to 21 years prior (7). In addition, workers in the erstwhile radium dial painting industry had been reported to have an increased risk of sinonasal SCC. The radium was believed to have been absorbed via the oral mucosa in workers who used to lick paintbrushes for painting the dials of watches. Sinonasal tumors were reported to arise after a median latent period of 34 years (7,27). The link between SCC and sinonasal papillomas is now well established. Sinonasal papillomas are benign epithelial neoplasms composed of well-differentiated, ciliated columnar or respiratory epithelium, with variable squamous metaplasia. Three types are identified: inverted papilloma, columnar cell (oncocytic Schneiderian) papilloma, and exophytic papilloma (28). Inverted papilloma is characterized by its endophytic or “inverted” growth pattern into the underlying
Squamous cell carcinoma (SCC) of the nose and paranasal sinuses is rare. Large-scale experience in management of this tumor is not available from a single institution. The relevance of most reported outcome data is further diluted by inclusion of tumors of various histologic types. Meaningful interpretation of results of treatment of sinonasal cancers is thus difficult because many published reports include other tumors such as adenocarcinoma and esthesioneuroblastoma, which have different biologic behavior and prognosis compared to SCC (1–4). A clear understanding of the biologic behavior of sinonasal SCC is therefore essential to treatment selection. Cancers arising in this anatomic area tend to remain relatively asymptomatic until late in their course, and thus tend to present at an advanced stage, by which time they have frequently extended to involve vital adjacent structures such as the orbit, skull base, or brain. Treatment of these tumors is therefore difficult and outcomes remain suboptimal in spite of advances in imaging and therapeutic techniques. The objective of this chapter is to review the epidemiology, etiology, pathology, natural course, and outcome of treatment of SCC arising in the nose and paranasal sinuses.
EPIDEMIOLOGY Malignant neoplasms of the nasal cavity and paranasal sinuses account for only 0.2% to 0.8% of all carcinomas, and only 2% to 3% of those in the upper aerodigestive tract (5,6). The incidence of cancers of the sinonasal tract in the United States and Europe is estimated at 0.3–1/100,000 per year (7,8). A higher incidence has been reported in Japan, Indonesia (9), and parts of Africa (10), with incidence rates of up to 2.6/100,000 per year reported in Japan (10). SCC is the most common histologic type (4,8,11,12) but precise incidence statistics for sinonasal SCC in the United States are not available. In Denmark, the annual incidence of SCC of the nose and paranasal sinuses has been estimated at 2.5 cases per million (11). The peak age incidence for SCC is in the sixth and seventh decades, however, adults of all ages may be affected (11). Sinonasal SCC occurring in a child is exceptionally rare. Males are more commonly affected than females by a factor of 1.5 to 2 (7,8).
ETIOLOGY A significantly increased risk of sinonasal SCC has been reported in nickel-refining workers (7,13,14). This risk appears 429
430
Sheahan et al.
stroma but with an intact basement membrane, and is the type that is usually associated with SCC, however, the columnar cell variety may also be associated with malignant transformation (29). The risk with exophytic papillomas is much smaller (28). The etiology of inverted papillomas is not clear. Human papillomavirus does appear to have a causal relationship with exophytic papilloma, but has been demonstrated in only a minority of cases of inverted papilloma (28,30,31). SCC of the nasal cavity and paranasal sinuses is more common in patients with a history of sinonasal papilloma. In fact, a recent study from Denmark suggested that 15% of cases of sinonasal SCC are associated with papillomas (32). The reported incidence of carcinoma associated with inverted papilloma has varied from as low as 2% (33) to as high as 53% (34), however, it is likely that incidences at the extremes of this range are due to either referral bias, papilloma misclassification, or failure to recognize small areas of SCC within inverted papillomas. More recent series report an SCC incidence of 5% to 26% (35–40). The carcinoma may be either synchronous (i.e., the diagnosis of SCC is established at the same time as that of the inverted papilloma), or metachronous (i.e., appearing at an area from where an inverted papilloma had previously been removed). Metachronous lesions usually appear at the same site as the inverted papilloma, and the mean time to their appearance is 52 months (41). On the basis of a review of personal series of 63 cases of inverted papilloma, along with a review of a further 3058 cases reported in the literature, Mirza estimated the incidence of synchronous SCC to be 7.1%, and of metachronous SCC to be 3.6% (41). While it is possible that some cases of metachronous SCC are due to failure to diagnose SCC, which was present at the time of the original resection, malignant transformation of inverted papillomas to SCC has been demonstrated histologically (42). Several other substances and occupations have been associated with sinonasal adenocarcinoma, but not with SCC. These include chromium, boot and shoe manufacturers in the leather tanning industry, textile workers, and isopropyl alcohol manufacturers (7).
PATHOLOGY AND NATURAL HISTORY SCC is the most commonly encountered epithelial malignancy of the sinonasal region and comprises 55% to 60% of all sinonasal tumors (8,10,43). It most commonly arises from the maxillary sinus (50–71%), followed by the nasal cavity (20–32%) and ethmoid sinus (10–15%) (8,10,43–45). Tumors of the sphenoid (2%) and frontal (≤1%) sinuses are exceedingly rare. Within the maxillary and ethmoid sinuses, males are more commonly affected than females, however, in the nasal cavity, females predominate over males (8). Most are poorly differentiated and keratinizing but about 20% are of the nonkeratinizing type (46). The histologic grade of these tumors does not predict outcome as reliably as their anatomic extent although poorly differentiated tumors are generally more aggressive in their course. Between 10% and 20% of paranasal sinus SCCs are very poorly differentiated (47). These so-called undifferentiated or anaplastic carcinomas are rapidly growing and produce early metastases. They are more evenly distributed throughout the maxillary sinus, nasal cavity, and ethmoid sinus, and consequently comprise a higher proportion of nasal and ethmoid carcinomas (8,10). They are said to be more common in females (8) and have a poorer prognosis than SCC (8,48). These tumors should be distinguished from other poorly differen-
tiated carcinomas, such as sinonasal undifferentiated carcinoma and malignant melanoma. A well-differentiated but nonkeratinizing variant is occasionally seen (2–11%). Names used to describe this variety include transitional cell carcinoma, cylindrical cell carcinoma, and Ringertz carcinoma (46). It is characterized by a plexiform or ribbon-like growth pattern and invades the underlying tissue with a smooth, well-delineated border. It is evenly distributed in all sinonasal locations (8,49). This variant is more common in males, in whom it occurs at a younger age than in females (8) and has a better prognosis than SCC (8). Rare variants of nasal and paranasal sinus SCC include papillary SCC, which is an exophytic carcinoma with a papillary configuration composed of thin fingers of tumor surrounding fibrovascular cores (46); verrucous carcinoma, which is an extremely well-differentiated lesion associated with minimal invasiveness (50); basaloid SCC, which consists of predominantly pleomorphic, basaloid-appearing cells, and is associated with a dismal prognosis (51); spindle cell carcinoma, an aggressive tumor with carcinomatous and spindle cell components (35), and adenosquamous carcinoma, a tumor containing areas of squamous carcinoma and areas of glandular differentiation, which is generally also considered to be highly aggressive (52). Small tumors contained within the sinus of origin are generally asymptomatic, thus presentation with early stage cancers is unusual. As the tumor enlarges, it comes into contact with the bony walls of the sinus and quickly destroys the bone, spreading into adjacent sinuses and other nearby structures, thence giving rise to symptoms and signs. Carcinomas of the maxillary sinus spread superiorly into the orbit, inferiorly into the hard palate and lower alveolus, medially into the nasal cavity, and anteriorly into the subcutaneous tissues of the cheek. More advanced tumors demonstrate posterior extension into the pterygomaxillary fissure, pterygoid plates, and infratemporal fossa, as well as superior extension up to the skull base and anterior spread through the cheek skin. Ethmoid tumors usually show early spread laterally into the orbit, superiorly through the cribriform plate into the anterior cranial fossa, and inferomedially to the nasal cavity and nasal septum. The lamina papyracea and cribriform plates are quickly destroyed; however, the periorbita and dura are more resistant to tumor invasion. Spread of tumor into the maxillary antrum is also commonly seen. Posterior spread into the sphenoid sinus and nasopharynx occurs later but spread to the frontal sinus is less common. Cancers of the nasal cavity most commonly (85%) arise from the lateral nasal wall. Less commonly, they may arise from the nasal septum, the nasal floor, or nasal vestibule (53). Tumors arising within the nasal cavity quickly expand to fill the nasal cavity. From there, extension occurs laterally into the ethmoid and maxillary sinuses, posteriorly into the nasopharynx, and superiorly to the skull base. Carcinomas arising from the nasal vestibule commonly involve the anterior septum and columella. From there, tumors have a propensity for aggressive spread along the periosteum of the premaxilla and maxilla. SCC of the vestibule is unusual in that it has been reported to demonstrate a more indolent natural course (54,55). Ohngren’s line (Fig. 1) represents an imaginary plane running from the medial canthus of the eye to the angle of the mandible, which has traditionally been used to separate sinonasal tumors into those with good and bad prognosis based on their location. Tumors situated below, medial, and anterior to this plane cause early symptoms, thus leading
Chapter 29: Squamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses
431
Figure 1 Ohngren’s line.
to diagnosis at an early stage, are easier to control surgically and are thus associated with a better prognosis. On the other hand, tumors situated above, lateral, and posterior to the plane present late due to a lack of early symptoms, have a higher incidence of invasion of the orbit, intracranial compartment, and infratemporal fossa, are more difficult to remove en bloc, and are therefore associated with a worse prognosis (56). Ohngren’s line formed the basis for the initial version of the UICC staging system for maxillary sinus carcinoma. Lymphatic drainage from the anterior part of the nose and paranasal sinuses is by means of facial lymphatic vessels into cervical lymph nodes at levels I and II. Drainage from the posterior part of the nose and paranasal sinuses is to retropharyngeal nodes, and thence to upper deep cervical nodes (57–59). Regional lymph node metastasis at the time of diagnosis is generally less common with SCC of the nose and paranasal sinuses than with SCC at other head and neck primary sites, with reported incidences ranging from 4% to 16% (60–68). There are little data on the incidence of occult disease in the clinically N0 neck since elective neck dissection is not routinely practiced for these tumors. Distant metastases at the time of presentation are rare and generally only seen in patients with regional disease (66).
with proptosis secondary to invasion into the orbit. A mass at the medial canthus, diplopia, and orbital pain may also be present. Paralysis of medial gaze secondary to invasion of the medial rectus muscle occurs later. Presenting symptoms of maxillary tumors include a facial mass, a palatal or upper alveolus mass, loosening of teeth or denture problems, facial or dental pain, or proptosis. Epiphora is caused by tumor invading or obstructing the nasolacrimal duct. Anesthesia of the cheek, upper lip, and upper teeth signifies involvement of the infraorbital nerve. Tumors of the frontal sinus usually present with a mass above the eye and symptoms of obstructive sinusitis. Signs indicative of more advanced tumors of the paranasal sinuses include trismus (invasion of the masticator muscles), numbness of the chin (mandibular nerve invasion at the foramen ovale), visual loss (secondary to optic nerve invasion), and cervical lymphadenopathy. The new appearance of any unilateral nasal symptoms, especially in an adult patient should always prompt careful endoscopic examination of the nasal cavity to rule out a tumor. Proptosis; external swelling of the face, periorbital region, nose, palate, or upper gum; unilateral epiphora; and facial numbness are all signs that should be regarded as highly suspicious of a tumor (49,69), and should always be promptly investigated by a CT scan.
SYMPTOMS AND SIGNS
WORKUP AND ASSESSMENT
Squamous carcinomas of the nose and paranasal sinuses generally do not produce any symptoms or signs until they have expanded to a significant size and/or have extended outside the bony confines of the sinus cavity. Nasal tumors most commonly present with unilateral nasal obstruction, rhinorrhea, or epistaxis. Bleeding may take the form of blood-stained nasal secretions. As the tumor enlarges, it may present into the nostril and/or cause external nasal deformity. Occasionally, patients may present with clear rhinorrhea secondary to cerebrospinal fluid leakage. Ethmoid tumors usually present
Workup of patients with sinonasal SCC should include a full history and physical examination, relevant laboratory studies, imaging studies, tissue diagnosis, and appropriate consultations from other specialists.
Imaging Radiologic imaging studies are an essential component in the diagnosis, staging, and follow-up of sinonasal malignancies. Computed tomography (CT) scan gives a good initial overview of the tumor’s location with excellent bone detail.
432
Sheahan et al.
Because the paranasal sinuses and nasal cavity are mucosal lined bony chambers, CT is helpful in determining whether a tumor remains confined within these natural boundaries or has eroded through the surrounding bone. CT provides details of the extent of local bone invasion, and is particularly useful in assessing the lamina papyracea, orbital floor, ethmoid roof, cribriform plate, pterygoid plates, hard palate, and skull base (70). The clinician should be aware that the brain, meninges, orbit, and facial soft tissues are inadequately evaluated when a bone algorithm and bone windows are used; when evaluating cancers in this region, soft tissue windows are also essential. However, magnetic resonance imaging (MRI) is preferred for accurate soft tissue details, so that most patients benefit from both CT and MRI. In comparison to CT, MRI allows a better distinction of tumor from adjacent soft tissue. MRI is particularly useful for determining invasion of the orbital contents, dura, brain, cavernous sinus, and infratemporal fossa (70). MRI may also be better for assessing carotid artery invasion, and newer techniques such as MRA permit intraluminal carotid assessment without the associated risks of direct angiography. MRI also differentiates fluid collection secondary to an obstructed sinus from tumor. CT and MRI therefore complement one another in the assessment of sinonasal tumors. CT provides excellent bone detail, while MRI offers better soft tissue imaging. In assessing CT and MRI scans, both coronal and axial views should be studied. Coronal views are particularly helpful in assessing invasion of the cribriform plate, lamina papyracea, and palate; while axial views are helpful in delineating posterior extension into the sphenoid sinus, orbital apex, pterygoid plates, pterygomaxillary fissure, or infratemporal fossa, and anterior extension into the cheek. In cases where tumor extension is present into the infratemporal fossa, coronal views allow assessment of the foramen ovale. The hallmark of sinonasal carcinomas on CT scans is bone destruction, which is seen in approximately 80% of scans at initial presentation (71). The tumor itself is usually of soft tissue density. Intravenous contrast causes tumor enhancement, however, inflamed mucosa may enhance similarly. Because of this, differentiation between tumor, mucosal thickening, and obstructed secretions may be difficult on CT scan. This distinction is greatly facilitated by MRI. Malignant tumors tend to be of intermediate signal on T2weighted image (72). In contrast, inflamed mucosa, retained secretions, and benign polyps generally have a high signal on T2-weighted images. In addition, even in cases where the secretions become increasingly inspissated and the signal intensity on T2-weighted scans decreases, the tumor can usually still be distinguished by its typical heterogeneity, in contrast to the smooth homogenous appearance of secretions (72). It should be noted that some schwannomas, minor salivary gland tumors, and inverted papillomas may also be bright on T2-weighted imaging. With gadolinium injection on T1-weighted images, tumors enhance less intensely than inflamed mucosa, while secretions do not enhance. Benign tumors extending intracranially tend to be more heterogeneous on MRI than malignant tumors (73). Mucoceles may be distinguished from tumors by peripheral enhancement (74). In addition to the above, imaging of the neck, using either CT or MRI, should be performed to assess for regional metastases. Chest CT and/or positron emission tomography (PET scan) may be performed to rule out distant metastases, however, in the absence of cervical metastases, distant metastases are very unlikely.
Tissue Diagnosis Biopsy Once the site and extent of the tumor has been identified, tissue diagnosis is required. A fundamental principle should be to obtain representative tissue by the least invasive method possible. Avoiding an open procedure is advantageous in preventing (i) the disturbance of intact anatomic structures and boundaries, (ii) possible tumor contamination of normal tissues, and (iii) disturbance of the tumor’s location and obscuration of its margins, making future localization and surgical treatment significantly more difficult. An optimal procedure for biopsy of sinonasal malignancies is through an endoscopic approach through the nares. This approach offers several advantages, including excellent visualization, low morbidity, and minimal alteration of the tumor and its surrounding structures. Even small, lateral tumors within the maxillary sinus may be accessible with the creation of a middle meatal antrostomy, visualization with a 45 degree or 70 degree endoscope, and biopsy using a long curved giraffe instrument. If the tumor presents itself at the nasal vestibule, biopsy in the office may be considered, however, in general, biopsy in the operating room is preferred as this allows the surgeon to deal with any bleeding that may arise. It is important to ensure by clinical and radiologic examination that the mass is neither contiguous with the cerebrospinal fluid space nor highly vascular. If the mass compresses easily or appears vascular then further imaging should be obtained prior to biopsy. In rare cases where a maxillary sinus tumor is not accessible transnasally with the endoscope, a canine fossa puncture can be combined with endoscopic visualization and biopsy. With the availability of endoscopic techniques, Caldwell– Luc antrostomy is rarely necessary. Open biopsy should be avoided at all costs as this violates tissue planes in the subcutaneous tissue and skin, which will then have to be sacrificed with a margin of normal skin at the time of definitive surgical resection.
Review of Histology Review of the biopsy slides by an experienced histopathologist is strongly advised in patients with sinonasal carcinoma. As discussed previously, the histologic type of the tumor is a powerful predictor of outcome and may also influence treatment selection. Accurate typing of tumor histology is therefore crucial prior to commencing treatment, especially since it has been recognized that up to one in five cases have the diagnosis modified after expert review (75).
Multidisciplinary Consultation The care of the patient with sinonasal SCC requires a multidisciplinary approach. In addition to the head and neck surgeon, consultations may be necessary from (i) neurosurgery, if the patient is going to require a craniofacial resection; (ii) plastic/reconstructive surgery, if the patient is going to require a free flap to reconstruct the surgical defect; (iii) prosthodontics, if the patient is going to require a dental obturator or other prostheses; (iv) radiation oncology; (v) medical oncology, if chemoradiation therapy is a consideration; (vi) internal medicine, if the patient has medical conditions that need to be optimized prior to surgery; (vii) ophthalmology, if the tumor has affected the eye or if treatment has the potential to impact visual function.
Chapter 29: Squamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses Table 1 AJCC/UICC Classification of Tumors of Maxillary Sinus
Table 3
Tx T0 Tis T1
N0 N1
T2
T3
T4a
T4b
Primary tumor cannot be assessed No evidence of primary tumor Carcinoma in situ Tumor limited to the antral mucosa with no erosion or destruction of bone Tumor causing bone erosion or destruction, including extension into the hard palate and/or middle meatus, but excluding extension to posterior wall of maxillary sinus or pterygoid plates Tumor invades any of the following: bone of posterior wall of maxillary sinus, subcutaneous tissues, floor or medial wall of orbit, pterygoid fossa, or ethmoid sinuses Tumor invades anterior orbital contents, skin of cheek, pterygoid plates, infratemporal fossa, cribriform plate, frontal or sphenoid sinuses Tumor invades any of the following: orbital apex, dura, brain, middle cranial fossa, cranial nerves other than maxillary division of the trigeminal nerve, nasopharynx and/or clivus
STAGING Tumors of the maxillary sinus, nasal cavity, and ethmoid sinus are staged according to the TNM system of the UICC/AJCC. The latest edition of this was published in 2002 (76). The staging system for maxillary sinus tumors is given inTable 1, and the staging system for tumors of the nasal cavity and ethmoid sinus is shown in Table 2. Currently, no staging system exists for tumors arising primarily in the frontal or sphenoid sinus. Neck staging is the same as for other head and neck sites (Table 3). Stage grouping is given in Table 4. Of note, for maxillary sinus carcinomas, extension into the infratemporal fossa, pterygoid plates, or skin of cheek, is now classified as T4a, and not as T3, as had been the case with the 1997 classification.
PROGNOSTIC FACTORS The following factors have consistently been shown to be significant prognostic factors for local control and survival in patients with sinonasal SCC: T stage (2,77–79), N stage (65,66,68,79,80), intracranial extension (2,80,81), dural invasion, orbital invasion (2,79), adverse tumor histology (80). Other adverse factors include advanced age (68,79). SCC asTable 2 AJCC/UICC Classification of Tumors of Nasal Cavity or Ethmoid Sinus Tx T0 Tis T1 T2
T3 T4a
T4b
Primary tumor cannot be assessed No evidence of primary tumor Carcinoma in situ Tumor restricted to any one subsitea , with or without bony invasion Tumor invading two subsites in a single region or extending to involve an adjacent region within the nasoethmoidal complex, with or without bony invasion Tumor extends to invade the medial wall or floor of the orbit, maxillary sinus, palate, or cribriform plate Tumor invades any of the following: anterior orbital contents, skin of nose or cheek, minimal extension to anterior cranial fossa, pterygoid plates, frontal or sphenoid sinuses Tumor invades any of the following: orbital apex, dura, brain, middle cranial fossa, cranial nerves other than maxillary division of the trigeminal nerve, nasopharynx and/or clivus
a Subsites are: right ethmoid sinus, left ethmoid sinus, nasal septum, nasal floor, nasal lateral wall, nasal vestibule.
N2a N2b N2c N3
433
AJCC/UICC Classification of the Neck Nodes
No regional lymph node metastasis Metastasis in single ipsilateral lymph node, 3 cm or less in greatest dimension Metastasis in single ipsilateral lymph node, greater than 3 cm but no more than 6 cm in greatest dimension Metastasis in multiple ipsilateral lymph nodes, none more than 6 cm in greatest dimension Metastasis in multiple contralateral or bilateral lymph nodes, none more than 6 cm in greatest dimension Metastasis in lymph node greater than 6 cm in greatest dimension
sociated with inverted papilloma has been associated with a better prognosis (82). Cervical metastasis is associated with a particularly poor prognosis, with reported 5-year survivals of no better than 10% to 15% (63,65,68). The most common cause of treatment failure in sinonasal SCC is local recurrence. Patients with local recurrence will usually die of disease. Hence, optimizing local control is of paramount importance in sinonasal SCC. Regional failure in cervical lymph nodes, even without elective neck treatment, is uncommon. Neck failure has been shown to be significantly associated with nodal stage (2,81). Among series containing only patients with SCC, distant metastases are reported in 11% to 13% (77,83,84). Factors associated with increased risk of developing distant metastases include high T stage (2), nodal metastases (81), intracranial invasion (2,81), and orbital invasion (81).
TREATMENT A wide variety of management approaches to sinonasal SCC exists. Most of these approaches involve some combination of surgery, radiotherapy, and/or chemotherapy. It is notable that there are no randomized trials comparing outcomes between the different treatment types. The treatment options for sinonasal SCC can be summarized as follows: (i) primary surgery, usually combined with postoperative radiotherapy; (ii) primary radiotherapy, reserving surgery for salvage of persistent disease; (iii) neoadjuvant chemotherapy, followed by various combinations of further chemotherapy, radiotherapy and/or surgery; (iv) radiotherapy, with or without neoadjuvant or concurrent chemotherapy, combined with “conservative surgery.” Comparing outcomes of the various treatment modalities for sinonasal SCC is problematic for the following reasons: (i) Sinonasal SCC is rare, so few centers have contemporary experience with a large volume of cases. (ii) Owing to the complex anatomy of the paranasal sinuses and surrounding structures, sinonasal SCCs comprise a heterogenous group of tumors with differing locations and extents of invasion. (iii) Most of the published series include a substantial number of tumors with histologies other than SCC. These other histologies, including adenocarcinoma and esthesioneuroblastoma, have very different biologic behavior and different Table 4 Stage 1 Stage 2 Stage 3 Stage 4
AJCC/UICC Stage Grouping for Sinonasal Carcinoma T1N0M0 T2N0M0 T3N0M0, or T1–T3, N1, M0 T4 or T1–T3, N2–N3 or M1
434
Sheahan et al.
implications for prognosis (1–4). (iv) In many series, patients with large advanced tumors, considered to be unresectable, are over-represented in the radiotherapy alone group, thus introducing a considerable bias when making comparisons between patients who did and did not undergo surgery (66). Prior to making a decision regarding treatment, a thorough assessment of the patient and tumor is essential. Particular points to consider include the following: 1. Tumor involvement of the cribriform plate, +/– intracranial extension. If this is present, then a craniofacial resection will be required to ensure adequate clearance. 2. Tumor invasion of the orbit. If this is present, then orbital exenteration may be required (see below). 3. Tumor invasion of the palate. If the palate or upper alveolus is involved, then infrastructure maxillectomy will be required. The patient will need to see a prosthodontist prior to surgery in order to have a dental obturator prefabricated for immediate fitting following tumor resection in order to close the palatal defect, unless the surgical defect is to be reconstructed using a free flap. 4. Extension into the sphenoid sinus. Tumor in the sphenoid sinus may also invade the internal carotid artery and cavernous sinus, as well as the optic nerve. These findings may render the tumor inoperable (see below). 5. Posterior extension into the masticator space, +/– involvement of the foramen ovale and middle cranial fossa. 6. Involvement of the skin. If skin is resected, the resulting defect may be very difficult to repair using local flaps. Tenuous repairs are at high risk of breaking down during postoperative radiotherapy, leading to a facial fistula. This is a particular problem near the medial infraorbital rim where the skin is usually very thin. Therefore, a free flap will be necessary in most patients that require resection of the skin. 7. Presence of cervical metastases. Patients with cervical metastases have a dismal prognosis; this should be borne in mind prior to proceeding with surgery if resection of the primary tumor is likely to entail significant morbidity. 8. General medical condition of the patient, and ability to withstand a prolonged surgery, including possible craniotomy and free flap reconstruction.
illectomy with complete ethmoidectomy +/– sphenoid exenteration; suprastructure maxillectomy +/– orbital exenteration; infrastructure maxillectomy; or total maxillectomy +/– orbital exenteration, resection of the pterygoid plates, and masticator space dissection. Resection of the skin of the cheek, nasal septum, and contralateral ethmoid and sphenoid sinuses may also be required. Our preferred approach is via a modified Weber–Ferguson incision respecting the nasal subunits, with or without subciliary (Fig. 2) or Lynch extension (Fig. 3). This may be combined with a bicoronal incision for craniofacial resection as required. For more limited infrastructure maxillectomies, a peroral or midfacial degloving approach (Fig. 4) may also be used. Medial maxillectomy combined with ethmoidectomy (Fig. 5) is indicated for tumors primarily arising in the ethmoid sinuses or nasal cavity with minimal invasion of the maxillary antrum and no invasion of the floor of the nose. The extent of surgery will vary with the location of the disease; however, in general this involves en bloc resection of the medial part of the maxilla from the infraorbital rim to the lower part of the piriform aperture, the inferior turbinate, the lacrimal bone, the lamina papyracea, and the anterior ethmoid air cells. In most cases, it should be possible to preserve the infraorbital nerve, as well as the roots of the upper teeth. The resection usually includes a variable proportion of the frontal process of the maxilla and the nasal bone. Tumors of the ethmoid sinuses that have suspected or obvious orbital invasion require discussion regarding orbital exenteration (see below). Access for medial maxillectomy is usually by means of a lateral rhinotomy incision with Lynch extension. Upper lip split should not be necessary. The anterior ethmoid artery is an important surgical landmark, which enters the orbit through or just below the frontoethmoidal suture line
In the discussion that follows, emphasis has been placed on studies reporting only on patients with SCC rather than those that included patients of various other histologies.
Surgery Surgery with postoperative radiotherapy continues to be the mainstay of treatment for sinonasal SCC in most centers throughout the world. Careful preoperative workup is essential to assess the resectability of these tumors, as well as the surgical approach. Surgical extirpation of tumors of the nose and paranasal sinuses can be challenging not just from the technical standpoint but also because the tumor and its treatment is likely to result in significant functional and cosmetic morbidity. Preoperative assessment of these tumors should thus take into consideration the site of origin and extension of the tumor, and the performance status of the patient, as the extent of resection in many patients may mandate a craniotomy, and/or reconstruction using a free flap. The ideal surgical procedure would be an en bloc resection of the entire tumor with negative microscopic margins. However, in practice, this is sometimes difficult to achieve. The type of surgery will depend on the location and extent of invasion by the tumor. Surgery may involve medial max-
Figure 2
Weber–Ferguson incision with subciliary extension.
Chapter 29: Squamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses
Figure 5
Figure 3 Weber–Fergusion incision with Lynch extension
roughly 14 to 18 mm posterior to the anterior lacrimal crest. This marks the junction between the lamina papyracea and skull base and so guides the upper limit of resection within the orbit. The posterior ethmoid artery is roughly 10 mm posterior to the anterior ethmoid artery and is a useful landmark for the optic nerve, which enters the orbit 4 to 7 mm posterior, but may be as close as 2 mm lateral (Fig. 6). The inferior turbinate is typically resected en bloc with the rest of the specimen; however, the middle turbinate is usually attached to the skull base and so has to be removed separately.
435
Medial maxillectomy.
For tumors that reach the level of the cribriform plate, craniofacial resection should be performed. This allows en bloc resection of the roof of the ethmoid and cribriform plate as well as the middle turbinate with the specimen. For tumors that arise in the nasal cavity, resection of the nasal septum is often necessary. In such cases, the anterior part of the septum should be preserved if possible in order to maintain support for the tip of the nose. Depending on the posterior extent of the tumor, exenteration of the posterior ethmoid and sphenoid sinus may be required. This should be performed to the level of the skull base, which is flat and relatively easy to identify in this region, however, owing to the proximity of such vital structures as the optic nerve and internal carotid
Anterior ethmoidal a. Ethmoid sinuses
Posterior ethmoidal a.
Optic nerve
Figure 4 Midfacial degloving incision.
Figure 6
Orbit with AEA, PEA, and optic nerve.
436
Sheahan et al.
artery, it is usually difficult to achieve monobloc resection of tumor involving these areas. For tumors that invade intracranially but remain extradural, craniofacial resection should be performed. Craniofacial resection also facilitates satisfactory tumor removal from the posterior ethmoid and sphenoid sinuses. Of note, the dura provides a good barrier against tumor invasion. In cases where the tumor erodes the cribriform plate and extends intracranially, then the dura should be resected and reconstructed in order to ensure clear margins in this area. Gross dural invasion may also be resected and reconstructed. Brain invasion with SCC is generally considered a contraindication to tumor resection since outcomes are dismal even in selected patients who are amenable to craniofacial resection (1). Suprastructure maxillectomy (Fig. 7) involves resection of the superior part of the maxilla, but preserving the hard palate and lower alveolus. It is indicated for ethmoid tumors with more extensive invasion of the maxillary antrum, or for maxillary tumors that do not involve the floor of the maxillary sinus or roots of the upper teeth. Typically, the infraorbital nerve and most of the orbital floor are removed. Depending on the extent of orbital invasion, orbital exenteration may also be performed. Suprastructure maxillectomy may also be combined with ethmoidectomy with or without craniofacial resection, removal of the posterior wall of the maxillary antrum, as well as variable removal of the pterygoid plates and contents of the pterygomaxillary fissure and pterygopalatine fossa. Because most of the orbital floor is removed, one of the main problems with suprastructure maxillectomy is postoperative support of orbital contents in cases where the globe has been preserved. In cases where the orbital periosteum and medial canthal ligament are fully preserved, then these structures may provide adequate support to suspend the or-
Figure 7
Suprastructure maxillectomy.
Figure 8 Infrastracture maxillectomy.
bital contents postoperatively. However, partial or complete removal of either or both of these may necessitate further measures to provide support. Possible reconstructive options in this situation include free fascia lata grafts, vascularized temporalis myofascial flaps, vascularized calvarial bone flaps (85), and free flaps. However, in practice, it is often difficult for pedicled flaps to fully bridge the defect, while access for free flap vessels is also difficult (86). Free bone grafts are not always the best option, as most of these patients will require postoperative radiotherapy, which will prevent neovascularization of the bone graft. Thus, such patients are at high risk of subsequently developing osteoradionecrosis. Similarly, in cases where the orbital floor is reconstructed using prosthetic materials, radiotherapy leads to a high incidence of implant exposure and infection (87). Infrastructure maxillectomy (Fig. 8) is indicated for tumors arising on the upper alveolus, hard palate, or lower part of the maxillary sinus that do not involve the ethmoid or the roof of the maxillary sinus. The resection includes part or all of the upper alveolus and hard palate. The floor of the orbit is spared. Depending on the extent of resection, the operation may be performed via a peroral, midfacial degloving, or upper lip-split approach. Reconstruction of the hard palate and upper alveolus is most easily accomplished using a prosthetic dental obturator. Although split-thickness skin grafts are frequently used on the inner surface of the cheek flap, our practice is to avoid using these as superior granulation and mucosalization of the cavity is achieved without them. Total maxillectomy (Fig. 9) is indicated for larger tumors, which involve most of or fill the maxillary sinus. Resection may also involve ethmoidectomy, orbital exenteration, and/or removal of the pterygoid plates and contents of the pterygomaxillary fissure and masticator space. The usual approach is via a modified Weber–Ferguson incision with subciliary extension. Care must be taken raising the
Chapter 29: Squamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses
437
chiasm involvement, bilateral cavernous sinus involvement, or internal carotid artery involvement (1,63). The results for surgery alone for sinonasal carcinoma are generally disappointing, with typical 5-year local control rates of 40% (66) and 5-year overall survival rates of 20% (65). Most centers now offer postoperative radiotherapy to all but the earliest stage tumors. Using combined treatment, typical 5-year local control rates range from 49% to 67% (65,81,88), with 5-year disease-specific survival of 64% reported (60), and reported 5-year overall survival ranging from 42% to 66% (81,89).
Radiotherapy
Figure 9
Total maxillectomy.
cheek flap as tumor may easily erode through the thin bone of the canine fossa to involve the subcutaneous tissues. Medially, the incisor teeth arising from the premaxilla may be preserved if this area is not involved by tumor. Laterally, the maxilla may be separated from the zygoma at the zygomatomaxillary suture, preserving the prominence of the cheek. However, for more extensive tumors, it may be necessary to divide the frontal process of the zygoma and the zygomatic arch in order to allow removal of the entire infraorbital rim. For tumors invading through the posterior wall of the maxillary sinus, the pterygoid plates should also be removed. This may be achieved by insinuating an osteotome behind the hard palate into the pterygoid fossa to fracture the top of the pterygoid plates from the skull base, and using a Mayo scissor to divide the pterygoid muscles from the lateral pterygoid plate. Alternatively, improved exposure of the region of the pterygoids may be achieved by either dividing the insertion of the temporalis muscle from the coronoid process of the mandible and then removing the coronoid process or by performance of a mandibular swing (60). Failure to mobilize the pterygoid plates will usually result in a fracture across the back wall of the maxillary sinus when the specimen is removed, necessitating piecemeal removal of tumor in the pterygomaxillary fissure and masticator space. Total maxillectomy usually results in a large defect, which requires reconstruction for both functional and cosmetic reasons. Reconstruction may be either with a prosthetic obturator or a free flap. Advantages of a free flap include the provision of support for the orbital contents, as well as avoidance of a large cavity, which will require fastidious irrigation and cleaning. In this situation, the best option is usually a rectus abdominis free flap, which provides a large volume, as well as epithelial surfaces, which may be used intraorally and/or intranasally (86). Contraindications to surgery are not universally accepted; however, they generally include the following: presence of distant metastases, presence of brain invasion or extensive intracranial invasion, bilateral optic nerve or optic
Substantial uncertainty surrounds the optimal radiation volumes and techniques for paranasal SCC. However, the results for radiotherapy alone in these tumors have been disappointing, with reported 5-year local control rates ranging from 14% to 53% (53,65,68,88,90–93) and 5-year survival rates ranging from 0–16% (43,65,80,88,92–93). The drawbacks of radiotherapy alone as treatment for paranasal sinus SCC are not only due to lack of efficacy, but also to the considerable potential for adverse effects to nearby structures such as the retina, optic nerves, optic chiasm, and frontal lobes. The incidence of serious visual complications with conventional radiotherapy has been reported to range between 16% to 66% (79,94–96). Visual complications include severe keratitis, cataracts, chronic tearing, retinopathy, optic neuropathy, optic atrophy, and blindness. Unilateral blindness has been reported to occur in up to 20% to 35% of patients with sinonasal cancers undergoing radiotherapy, with bilateral blindness occurring in up to 6% to 10% (97–100). The risk to the eye would appear to be particularly high for ethmoid carcinomas (94,97,101), although high incidences of ipsilateral blindness (30%) have also been reported in patients with maxillary carcinomas (102,103). The most common scenario for radiation-induced blindness is radiation retinopathy of an eye irradiated to a high dose; however, contralateral blindness secondary to optic neuropathy has been reported to occur in 8% of cases (99). The risk of ocular complications appears to be related to the total dose of radiation, as well as to the treatment ports and treatment schedule. Parsons found retinopathy to occur in 100% of eyes receiving radiation doses of greater than 65 Gy, in 50% of eyes exposed to doses between 45 Gy and 55 Gy, while the risk was very low in eyes receiving less than 45 Gy (104). The risk of optic neuropathy appears to be increased in optic nerves receiving doses of 60 Gy or greater, in fractions of 1.9 Gy or larger (105). With modern radiotherapy techniques, acceptable maximum doses to vital optic structures are generally accepted to be 45 Gy to the retina, and 54 Gy to the optic nerve and chiasm (81,106). It has been suggested that hyperfractionation may reduce the risk of ocular complications (101). Other complications of radical radiotherapy for sinonasal carcinomas include brain necrosis (0–12%) (48,95,99,103), bone necrosis (0–8%) (2,80,92,99,103), hearing loss (2.5–8.5%) (2,3), hypopituitarism (4–5%) (99,103), trismus (5–12%) (3,92,103), facial fistula (2–3%) (92,103), saddle-nose deformities (95), diplopia (95), meningitis (48), and radiationinduced tumors (99). In addition, a study by Meyers et al. suggested that memory impairment may occur in up to 80% of patients treated with radiotherapy to the skull base by conventional techniques (107). One of the limiting factors of radiotherapy alone for sinonasal carcinoma is the high dose of radiation that is necessary, leading to an increased risk of bilateral blindness. The
438
Sheahan et al.
advantage of using combined surgery and radiotherapy is that a lower dose of radiation may be used, thus reducing the risk of bilateral blindness. In the University of Florida experience, bilateral blindness did not develop in any patient undergoing combined treatment; thus, the treatment policy at that institution was changed from one of irradiation alone, to one of surgical resection followed by radiotherapy (99). Recent evolutions in radiotherapy techniques may allow for improved efficacy and reduced adverse effects. Modern techniques such as intensity-modulated radiotherapy (IMRT) allow for very homogenous dose distributions with a sharp dose fall-off gradient, and thus an increased therapeutic ratio. Using these techniques, it is possible to irradiate the tumor with doses of up to 70 Gy, while keeping the maximum doses to the surrounding ocular structures to acceptable levels (95,106,108). In a review of 127 patients with sinonasal carcinomas of various histologies, treated over five decades, with conventional radiotherapy, 3-D conformal radiotherapy, and IMRT, Chen reported that while there was no difference in local control or survival according to the decade of treatment, a significant decrease in complications was seen. The incidence of late grade 3/4 ocular toxicity among patients treated with conventional radiotherapy, 3-D conformal radiotherapy, and IMRT was 20%, 9%, and 0%, respectively (95). Other authors also reported a low incidence of serious complications with 3-D conformal (2,109,110) and IMRT techniques (81,106), as well as with the use of proton beam irradiation (111). Duthoy prospectively recorded toxicity data of 39 patients with sinonasal carcinoma who were treated postoperatively with IMRT to a median dose of 70 Gy. Visual impairment developed in five (15%), two (6%) of whom had grade 3 impairment. There were two cases of brain necrosis (108). It should be noted that most patients in these studies underwent radiotherapy in a postoperative setting, and there are little data on the efficacy of IMRT as a primary treatment modality. Furthermore, radiation-induced optic neuropathy and retinopathy typically take 2 to 5 years to develop, thus the follow-up periods of some of these studies are relatively short (108,109). In the future, further dose reductions to normal tissue may be possible with the use of proton beam radiotherapy (112). When radiotherapy is combined with surgery, the optimal sequencing of these modalities has been debated. Dirix reporting on patients with various histologies, found local control and disease-free survival to be better in patients receiving postoperative compared to preoperative radiotherapy (2). Other authors have reported an improved survival in patients who received preoperative radiotherapy compared with those who received postoperative radiotherapy (113,114), however, a higher complication rate is reported among patients undergoing surgery after radiotherapy (1,113). In cases where patients with sinonasal carcinoma fail initial treatment with radiotherapy, the results of salvage surgery are disappointing. The present treatment policy at the University of Toronto consists of primary high-dose radiotherapy with curative intent to the equivalent of 70 Gy (63). Curran reported on 95 patients with sinonasal cancer of various histologies who failed initial treatment at the University of Toronto. Of these, 17 had distant metastases, 20 were deemed to have unresectable disease, and 24 were medically unfit or refused surgery at the time of the recurrence. Thus, 34 of the 95 patients proceeded to undergo salvage surgery. Within this group, the disease-specific 5-year survival was 47%, and the overall 5-year survival was 35%. The 5-year survival for patients with SCC was only 25%. Patients undergoing primary radiotherapy for sinonasal cancer in that
institution now undergo repeat imaging, examination under anesthesia, and biopsy of any suspicious areas 6 to 8 weeks after completion of radiotherapy, in an effort to detect residual or recurrent disease at an earlier stage (63,115). Of note, radiotherapy alone has been reported to be an effective treatment with good cosmetic outcomes for carcinomas of the nasal vestibule (54,116,117).
Chemotherapy Combined with Surgery and Radiotherapy In 1970, Sato et al. reported promising treatment outcomes in sinus carcinoma with a combination of necrotomy and radiotherapy concurrent with intra-arterial chemotherapy, for preventing functional and cosmetic loss caused by radical surgery. Complete response was achieved in 67% (118). Since then, several authors, mostly from Japan, have reported on the use of neoadjuvant chemotherapy prior to definitive treatment with surgery, radiotherapy, and/or chemoradiotherapy. The most common agents used are intra-arterial cisplatin (83,119) or 5-fluorouracil (5-FU) (61,67,83). These are administered by selectively introducing an intra-arterial catheter into the maxillary artery via the femoral artery or the superficial temporal artery (61,67). Tumor staining may be examined by CT angiography, and if insufficient, angiography of other arteries, including the facial (119) or ascending pharyngeal (120) may be added. The usual dose of cisplatin varies from 100 to 150 mg/m2 (119,120), given for 2 to 4 cycles (77,119,120). Disadvantages of intra-arterial chemotherapy include possible irregular distribution of the infused drugs due to the existence of alternate feeding vessels, and possible damage to branches of the external carotid artery, which may be required at a later date for free flap reconstruction of skull base defects. Thus, the use of systemic chemotherapy has been favored by other authors (121,122). Various combinations of chemotherapy, radiotherapy, and surgery have been reported. Samant reported on the use of concurrent intra-arterial chemotherapy combined with 50 Gy of radiotherapy, followed by surgery on 19 patients with sinonasal malignancy (14 with SCC). Five-year disease-free and overall survivals were both 53% (120). Lee reported overall survival of 72.7% on a subgroup analysis of 19 patients with stage III and IV paranasal sinus carcinoma (11 with SCC) who were treated with preoperative systemic chemotherapy with cisplatin and 5-FU, followed by radical surgery, followed by postoperative chemoradiotherapy using hydroxyurea and 5-FU, with a median radiation dose of 60 Gy (121). Konno reported a 5-year survival of 75% among 32 patients with maxillary SCC treated by concurrent radiotherapy (60 Gy) and intra-arterial 5-FU and cisplatin followed by radical surgery (67). There are little data regarding the outcome of patients treated with primary chemoradiotherapy without surgery. High progression-free rates have been reported even in patients with skull base invasion (123), although further validation of this initial observation is required in larger series. Among most reported series of patients undergoing treatment with primary chemoradiotherapy, surgery has been reported to be a significant predictor of improved outcome, although such series are biased by the inclusion of patients with unresectable disease in the nonsurgical group (77).
Combinations of Chemotherapy, Radiotherapy, and “Debulking” Surgery The combination of conservative “debulking” surgery with radical radiotherapy in order to avoid the morbidity of radical surgery has been reported. Debulking surgery has been
Chapter 29: Squamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses
defined as removal of all macroscopic tumor without additional removal of noninvolved bony structures. This removal is usually accomplished piecemeal, and may be performed via a lateral rhinotomy or, more commonly, by a sublabial approach (79,124). During the initial debulking surgery, a cavity may be established, which permits subsequent further debulking in the clinic (124). Alternatively, patients may undergo repeat debulking via an open approach (84,124). Using this approach, 5-year local control rates of 59% to 67% have been reported (79,124). Of note, Jansen reported that the combination of debulking surgery with radiotherapy offered a significant survival advantage when compared to radiotherapy alone (79). The absence of gross residual disease as judged by the surgeon after debulking has also been associated with a better outcome (124). The addition of neoadjuvant or concurrent chemotherapy to radiotherapy and debulking surgery has also been reported. Using these approaches, 5-year disease-free and overall survival rates of 88% and 72% to 76% respectively, have been reported (83,84). An interesting study conducted by Knegt et al. in 1985 described the use of surgical debulking and low-dose radiotherapy followed by repeated topical chemotherapy using 5-FU and necrotomy in patients with paranasal sinus cancer. The 5-year survival rate for SCC and undifferentiated carcinoma of the maxillary sinus was 52%. This protocol would appear to be particularly suited to adenocarcinomas. In 2001, the same group reported on this treatment for 62 patients with adenocarcinoma of the ethmoid sinus complex. Forty-nine patients had T3/T4 tumors, and 40% had anterior skull base involvement. The use of low-dose radiotherapy was omitted in the latter part of the study. Impressive 5- and 10-year disease-specific survival rates of 87% and 74% were reported (125). However, there have been little further data regarding the use of this treatment protocol for sinonasal SCC from other institutions. The obvious advantage of debulking surgery over conventional surgery is that it avoids the functional and cosmetic losses associated with more radical surgery, and avoids orbital exenteration. In addition, tumor volume is well known to be an important predictor of the radiocurability of a tumor. Reduction of the tumor volume by debulking surgery may thus enhance the likelihood of cure by radiotherapy (61). Kawashima found gross tumor volume after debulking surgery to be a more important predictor of local control than T-classification (61). Furthermore, radiotherapy dose of greater than 60 Gy was found to be an important predictor of local control in patients with small volumes of gross residual tumor. On the other hand, recent advances in free flap reconstructive surgery can now provide satisfactory restoration of masticatory and phonatory function after radical surgery, as well as good cosmetic results (86). Nibu suggested that the improved survival in patients treated in latter years at their institution may be in part due to their undergoing en bloc tumor resection rather than the piecemeal resections from earlier years (83). It should also be borne in mind that in studies reporting on the efficacy of debulking surgery, the presence of gross residual disease (122,124) was found to be a significant adverse prognostic indicator.
Comparison of Various Treatment Protocols Despite the numerous diverse combined treatment protocols, there are little data comparing the outcomes of any of these. Tiwari reported on 38 patients treated with curative intent for maxillary sinus SCC. Twenty-nine were treated by surgery
439
followed by radiotherapy, and had a 5-year survival rate of 64%. Nine were treated by sequential chemoradiotherapy, and had a 37% 2-year survival. The authors suggested that surgery followed by radiotherapy remains the treatment of choice for maxillary SCC (60). Shiga reported on 50 patients with maxillary sinus SCC. All received neoadjuvant intraarterial cisplatin. In 25 patients, this was followed by radical surgery with postoperative radiotherapy, while in the other 25, this was followed by concurrent chemoradiotherapy. There was no significant survival difference between the two groups, however, grade 3/4 toxicity was higher in the concurrent chemoradiotherapy group (119). Isobe found no significant difference between three different treatment protocols (77). The authors practice en bloc surgical resection of the tumor with appropriate reconstruction followed by adjuvant radiation with or without chemotherapy as the standard of care for surgically resectable tumors of the paranasal sinuses at their institution.
Management of the Orbit The specific indications for orbital preservation and exenteration have evolved over the past forty years and remain a controversial subject. Although in the 1950s, the orbit was almost routinely exenterated for any extension of maxillary sinus carcinoma toward the orbital floor, the emerging consensus is that the orbit can often be preserved without compromising overall survival or local control of disease. This approach has been complicated, however, by differing criteria that have been used to determine the indications for orbital preservation. Carrau examined 58 patients with bony orbital invasion by SCC of the sinonasal tract and found that 3-year survival was not affected by orbital preservation in the absence of orbital soft tissue invasion. The authors concluded that the orbit may be spared if the full thickness of the periorbita is not breached by tumor (126). McCary and Perry concluded that periorbital invasion does not necessarily indicate a need for orbital exenteration. They found that preoperative radiation therapy followed by intraoperative frozen section and selective resection of the involved periorbita may conserve the eye without a compromised outcome (127,128). Tumor extension through the periorbita does not necessarily condemn the eye to exenteration. Tiwari has noted that a thin fascial layer exists around the periorbital fat that is distinct from the periorbita and believes that invasion of this layer should determine the need for exenteration (129). Quatela has taken an even more aggressive approach by resecting intraorbital tumor with involved orbital fat and extraocular muscles off Tenon’s fascia surrounding the globe, and then preserving or “banking” the residual, nonfunctional globe in vivo (130). Care must be taken to avoid attempting orbital preservation at the potential cost of decreased local disease control and survival. Our approach is to resect involved periorbita and preserve the orbital contents in cases of periorbital involvement. Indications for orbital exenteration include the following: (i) invasion of orbital fat, (ii) invasion of extraocular muscles, (iii) invasion of bulbar conjunctiva or sclera, (iv) involvement of the orbital apex. Besides oncologic outcome, the other main point of contention with orbital preservation in sinonasal malignancy is functional outcome. Stern reported that only one-sixth of patients undergoing resection of the orbital floor retained useful function of the ipsilateral eye (131). Notably, no reconstructive efforts were made to repair lost orbital support. When efforts are made to reconstruct orbital floor defects,
440
Sheahan et al.
improved functional outcomes are seen (132). Among patients with sinonasal malignancy undergoing orbital preservation, between 86% and 91% of patients are reported to retain a useful functioning eye, however, 41% to 50% of these develop one or more ocular sequelae (132,133). The most common problem reported is abnormal globe position (enophthalmos or hypophthalmos). Abnormalities of globe position occur more frequently and are of greater severity in patients undergoing subtotal or total orbital floor resection, particularly if reconstructive efforts to support the globe are not employed. Only a minority of patients with abnormal globe position develop diplopia, and in most of these, diplopia is transient (132). Other ocular sequelae include epiphora and lid malposition (ectropion, canthal dystopia). The incidence of epiphora would appear to be decreased in patients undergoing lacrimal stenting (132). Of note, eye problems are more common when postoperative radiation therapy is administered (131,133).
Management of the Neck Clinically positive metastatic disease in the neck is generally managed with neck dissection, the type depending on the extent and location of the nodal metastases. Careful examination of parotid and facial lymph nodes should be performed prior to surgery as these nodes may occasionally also be involved. In addition, imaging should be performed to investigate the status of retropharyngeal nodes. Lymph nodes in level I, even if not clinically palpable, should always be included in the specimen in patients who have metastatic disease at other levels. Postoperative radiation therapy to the neck is indicated for multiple positive nodes, any single node >3 cm in size, or extracapsular spread. The management of the clinically negative neck remains controversial. Among patients with SCC who do not undergo elective neck treatment, the incidence of subsequent neck failure in the clinically N0 neck is generally reported to be between 5% to 14% (64,67,68,79,84). In patients who do develop regional failure, the results of salvage treatment are generally disappointing (64). Neck failure is also reported to be strongly associated with the development of distant metastases and decreased survival (64,68). Two widely quoted studies have reported incidences of neck failure as high as 29% to 33% (92,103), however, it should be noted that both of these studies included tumors of various histologies. Furthermore, the incidence of isolated neck failure in patients who were clinically N0 at the time of initial presentation in these studies was only 14% and 18% respectively. Of note, a recent systematic review and meta-analysis reported a weightedaverage incidence of neck failure for nasal cavity SCC of 18.1%; however, of the 23 studies reviewed, 20 reported exclusively on patients with SCC of the nasal vestibule or septum (134). Some studies have supported the use of elective neck irradiation for SCC of the maxillary sinus (64,89,103,135) and nasal cavity (134). Le reported that neck failure in his series was not predicted by primary tumor control, but was effectively prevented by elective neck irradiation (64). Jiang also found radiotherapy to effectively prevent neck failure (103). Despite this, given the low risk of neck conversion reported in most series, most institutions do not electively treat the clinically N0 neck. Exceptions may apply in cases where the sinonasal tumor encroaches upon areas of increased risk for lymphatic spread such as the nasopharynx or soft palate.
Management of Skin or Cartilage Invasion Cancers of the nasal vestibule and maxillary sinus and ethmoid sinus may commonly involve the skin of the cheek
or the nose. Even more common is for carcinomas of the maxillary sinus to break through the front wall of the maxillary sinus and involve the subcutaneous tissues, but not the skin itself. In such cases, care should be taken to raise a thin cheek flap, in order to leave an adequate margin of soft tissue over the tumor. When this is possible, skin resection is not necessary. On the other hand, when the skin is involved, then resection of the involved portion of skin is essential. An adequate margin of normal skin will need to be taken around the involved area. This usually creates a considerable soft tissue defect, which will require a free flap for reconstruction. In cases where there is also a large volume surgical defect, this is usually best accomplished using a rectus abdominis free flap. In cases where the volume of the defect is small, then a radial forearm free flap is better suited. Involvement of the skin and cartilages of the external nose is more problematic, as surgical resection will create a considerable cosmetic defect. On occasion, total rhinectomy, along with resection of the upper lip will be necessary. If possible, some or all of the nasal bones should be preserved in order to facilitate future placement of osseointegrated implants. Subsequent surgical reconstruction of the nose is extremely difficult, and the most satisfactory results are oftentimes obtained using a nasal prosthesis. If the upper lip is resected, this is reconstructed using advancement flaps, while a dental obturator is used to reconstruct the premaxilla. The surgeons should be wary of attempting to undertake immediate surgical reconstruction of the external nose unless they can be certain that the margins are negative. In many instances, delayed reconstruction may be a better option and this may need to be deferred until after adjuvant treatment is complete.
Role of Endoscopic Surgery Recent advances in endoscopic sinus surgery has led to increasing interest in the application of this technique to benign sinonasal and skull base tumors. More recently, the use of endoscopic surgery for malignant tumors has also been reported. Shipchandler reported on 11 patients with SCC of the nose or paranasal sinuses who were treated with endoscopic resection, combined with craniotomy in four cases. Eight patients received adjuvant chemotherapy, radiotherapy, or both. Two patients developed local recurrence and underwent repeat endoscopic resection. At a median followup of 31 months, 10 patients (91%) were alive with no evidence of disease (136). Buchmann et al. reported on 63 patients who underwent surgery for nasal/paranasal sinus cancer (26 with SCC). Of these, 27 were operated using a classic open approach, whereas 36 were operated using endoscopic techniques, either alone or in combination with a midfacial degloving approach or subfrontal craniotomy. There was no significant difference in survival outcomes. However, it is unclear which histologic tumor types underwent what type of surgery, and which cases received salvage treatment with surgery, radiotherapy, and chemotherapy (137). At the current time, caution should be exercised in treating patients with SCC of the paranasal sinuses with endoscopic resection since incomplete excision in inexperienced hands is liable to adversely impact outcomes even if the patient is subsequently subjected to more radical treatment.
OUTCOMES AND PROGNOSIS The prognosis of nasal cavity and paranasal sinus SCC continues to be suboptimal. Tumors generally present at an advanced stage. Many reported 5-year local control and
Chapter 29: Squamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses
survival rates are misleading as studies include patients with histologies other than SCC. In addition, most series include patients with tumors of all stages and all sites; however, by and large, the bulk of patients with SCC in any given series have advanced (T3/T4) primary tumors, with clinically negative (N0) necks. Among series reporting results only for patients with SCC after surgery with postoperative radiotherapy, typical 5year local control, disease-specific, and overall survival rates are 49% to 67% (65,81,88), 64% (60), and 42% to 66% (81,89) respectively. After combined treatment with chemotherapy, radiotherapy, and surgery, reported 5-year local control rates range from 59% to 88% (77,84,119,121,124), with 5-year disease-specific survival rates ranging from 52% to 75% (77,84,119,120,121), and 5-year overall survival rates ranging from 52% to 75% (67,77,83,119). Clearly, despite the varied treatment protocols, the results of treatment continue to be disappointing and novel approaches are needed to improve both oncologic and functional outcomes in these patients.
REFERENCES 1. Ganly I, Patel SG, Singh B, et al. Craniofacial resection for malignant paranasal sinus tumors: Report of an international collaborative study. Head Neck. 2005;27:575–584. 2. Dirix P, Nuyts S, Geussens Y, et al. Malignancies of the nasal cavity and paranasal sinuses: Long-term outcome with conventional or three-dimensional conformal radiotherapy. Int J Radiat Oncol Biol Phys. 2007;69:1042–1050. 3. Ozsaran Z, Yalman D, Baltalarli B, et al. Radiotherapy in maxillary sinus carcinomas: Evaluation of 79 cases. Rhinology. 2003;41:44–48. 4. Bhattacharyya N. Factors affecting survival in maxillary sinus cancer. J Oral Maxillofac Surg. 2003;9:1016–1021. 5. Osguthorpe JD. Sinus neoplasia. Arch Otolaryngol Head Neck Surg. 1994;120;19–25. 6. Ali S, Tiwari RM, Snow GB, et al. Incidence of squamous cell carcinoma of the head and neck: Report of 1000 cases. J Laryngol Otol. 1986;100:315–327. 7. Rousch GC. Epidemiology of cancer of the nose and paranasal sinuses: Current concepts. Head Neck Surg. 1979;2:3–11. 8. Robin PE, Powell DJ, Stansbie JM. Carcinoma of the nasal cavity and paranasal sinuses: Incidence and presentation of different histological types. Clin Otolaryngol. 1979;4:431–456. 9. Rifki N. Problems of paranasal sinus malignancy in Indonesia. ORL Indonesia. 1985;16:175–180. 10. Muir CS, Nectoux J. Descriptive epidemiology of malignant neoplasms of nose, nasal cavity, middle ear and accessory sinuses. Clin Otolaryngol. 1980;5:195–211. 11. Grau C, Jakobsen MH, Harbo G, et al. Sino-nasal carcinoma in Denmark 1982–1991— A nationwide survey. Acta Oncol. 2001;40:19–23. 12. Myers LL, Nussenbaum B, Bradford CR, et al. Paranasal sinus malignancies: An 18-year single institution experience. Laryngoscope. 2002;112:1964–1969. 13. Andrews P. Carcinomas of the nose and paranasal sinuses in former employees of a sinter plant at Copper Cliff, Ontario. J Otolaryngol. 1983;12:255–256. 14. Sunderman FW Jr, Morgan LG, Andersen A, et al. Histopathology of sinonasal and lung cancers in nickel refinery workers. Ann Clin Lab Sci. 1989;19:44–50. 15. Andersen A, Berge SR, Engeland A, et al. Exposure to nickel compounds and smoking in relation to incidence of lung and nasal cancer among nickel refinery workers. Occup Environ Med. 1996;53:708–713. 16. Hemelt M, Granstrom C, Hemminki K. Occupational risks for nasal cancer in Sweden. J Occup Environ Med. 2004;10:1033– 1040. 17. Luce D, Leclerc A, Morcet JF, et al. Occupational risk factors for sino-nasal cancer: A case-control study in France. Am J Ind Med. 1992;21:163–175.
441
18. Demers PA, Kogevinas M, Boffetta P, et al. Wood dust and sinonasal cancer: Pooled reanalysis of twelve case-control studies. Am J Ind Med. 1995;28:151–166. 19. Fukuda K, Shibata A. Exposure-response relationships between woodworking, smoking, or passive smoking, and squamous cell neoplasms of the maxillary sinus. Cancer Causes Control. 1990;1:165–168. 20. Hernberg S, Westerholm P, Schultz-Larsen K, et al. Nasal and sinonasal cancer. Connection with occupational exposures in Denmark, Finland, and Sweden. Scand J Work Enviro Health. 1983;9:315–326. 21. Luce D, Gerin M, Morcet JF, et al. Sinonasal cancer and occupational exposure to textile dust. Am J Ind Med. 1997;32:205–210. 22. Luce D, Leclerc A, Begin D, et al. Sinonasal cancer and occupational exposures: A pooled analysis of 12 case-control studies. Cancer Causes Control. 2002;13:147–157. 23. Hayes RB, Raatgever JW, de Bruyn A, et al. Cancer of the nasal cavity and paranasal sinuses, and formaldehyde exposure. Int J Cancer. 1986;37:487–492. 24. Brinton LA, Blot WJ, Becker JA, et al. A case-control study of cancers of the nasal cavity and paranasal sinuses. Am J Epidemiol. 1984;119:896–906. 25. ‘t Mannetje A, Kogevinas M, Luce D, et al. Sinonasal cancer, occupation, and tobacco smoking in European women and men. Am J Ind Med. 1999;36:101–107. 26. Hayes RB, Kardaun JW, de Bruyn A. Tobacco use and sinonasal cancer: A case-control study. Br J Cancer. 1987;56:843–846. 27. Littman MS, Kirsh IE, Keane AT. Radium-induced malignant tumors of the nose and paranasal sinuses. Am J Roentgenol. 1978;131:773–785. 28. von Buchwald C, Bradley PJ. Risks of malignancy in inverted papilloma of the nose and paranasal sinuses. Curr Opin Otolaryngol Head Neck Surg. 2007;15:95–98. 29. Kaufman MR, Brandwein MS, Lawson W. Sinonasal papillomas: Clinicopathologic review of 40 patients with inverted and oncocytic schneiderian papillomas. Laryngoscope. 2002;112:1372– 1377. 30. Buchwald C, Norrild B. Human papillomavirus in sinonasal papillomas and in normal nasal mucosa. Papillomavirus Report. 1997;8:99–104. 31. Weiner JS, Sherris D, Kasperbauer J. Relationship of human papillomavirus to Schneiderian papillomas. Laryngoscope. 1999;109:21–26. 32. Buchwald C, Lindeberg H, Pedersen BL, et al. Human papilloma virus and p53 expression in carcinomas associated with sinonasal papillomas: A Danish epidemiological study 1980– 1998. Laryngoscope. 2001;111:1104–1110. 33. Batsakis JG. The pathology of head and neck tumors: Nasal cavity and paranasal sinuses. Head Neck Surg. 1980;2:410– 419. 34. Yamaguchi KT, Shapshay SM, Incze JS, et al. Inverted papilloma and squamous cell carcinoma. J Otolaryngol. 1979;8:171– 178. 35. Lund VJ. Granulomatous diseases and tumors of the nose and paranasal sinuses. In: Kennedy DW, Bolger WE, Zinreich SJ, eds. Diseases of the Sinuses. Diagnosis and Management. London: BC Decker, 2001:85–106. 36. Philips PP, Gustafson RO, Facer GW. The clinical behavior of inverting papilloma of the nose and paranasal sinuses: Report of 112 cases with review of the literature. Laryngoscope. 1990;100:463–469. 37. Segal K, Atar E, Mor C, et al. Inverting papilloma of the nose and paranasal sinuses. Laryngoscope. 1986;4:394–398. 38. Smith O, Gullane PJ. Inverting papilloma of the nose: Analysis of 48 patients. J Otolaryngol. 1987;16:154–156. 39. Benninger MS, Roberts JK, Sebek BA. Inverted papillomas and associated squamous cell carcinomas. Otolaryngol Head Neck Surg. 1990;103:457–461. 40. Schlosser RJ, Mason JC, Gross CW. Aggressive endoscopic resection of inverted papilloma: An update. Otolaryngol Head Neck Surg. 2001;25:49–53. 41. Mirza S, Bradley PJ, Acharya A, et al. Sinonasal inverted papilloma: Recurrence, and synchronous and metachronous malignancy. J Laryngol Otol. 2007;121:857–864.
442
Sheahan et al.
42. Lawson W, Kaufman MR, Biller HF. Treatment outcomes in the management of inverted papilloma: An analysis of 160 cases. Laryngoscope. 2003;113:1548–1556. 43. Lewis JS, Castro EB. Cancer of the nose and paranasal sinuses. J Laryngol Otol. 1972;86:255–262. 44. Weinstein GS, Harvey R. Open surgical resection of sinus neoplasms. In: Kennedy DW, Bolger WE, Zinreich SJ, eds. Diseases of the Sinuses. Diagnosis and Management. London: BC Decker, 2001:405–422. 45. Spiro JD, Soo KC, Spiro RH. Squamous cell carcinoma of the nasal cavity and paranasal sinuses. Am J Surg. 1989;158:328– 332. 46. Pilch BZ, Bouquot J, Thompson LD. Squamous cell carcinoma. In: Barnes L, Eveson JW, Reichart P, Sidransky D, eds. World Health Organization Classification of Tumours. Lyon: IARC Press, 2000:15–17. 47. Weber AL, Stanton AC. Malignant tumors of the paranasal sinuses: Radiologic, clinical, and histopathologic evaluation of 200 cases. Head Neck Surg. 1984;3:761–776. 48. Uchida D, Shirato H, Onimaru R, et al. Long-term results of ethmoid squamous cell or undifferentiated carcinoma treated with radiotherapy with or without surgery. Cancer J. 2005;11:152– 156. 49. Panchal L, Vaideeswar P, Kathpal D, et al. Sino-nasal epithelial tumours: A pathological study of 69 cases. J Postgrad Med. 2005;51:30–35. 50. Daoud A, Lannigan FJ, McGlashan JA, et al. Verrucous carcinoma of the maxillary antrum. J Laryngol Otol. 1991;105:696– 699. 51. Wieneke JA, Thompson LD, Wenig BM. Basaloid squamous cell carcinoma of the sinonasal tract. Cancer. 1999;85:841–854. 52. Alos L, Castillo M, Nadal A, et al. Adenosquamous carcinoma of the head and neck: Criteria for diagnosis in a study of 12 cases. Histopathology. 2004;44:570–579. 53. Batsakis JG. Pathology of tumors of the nasal cavity and paranasal sinuses. In: Thawley SE, Panje WR, Batsakis JG, Lindberg RD, eds. Comprehensive Management of Head and Neck Tumors. Philadelphia: W.B. Saunders Co, 1999:522–539. 54. McCollough WM, Mendenhall NP, Parsons JT, et al. Radiotherapy alone for squamous cell carcinoma of the nasal vestibule: Management of the primary site and regional lymphatics. Int J Radiat Oncol Biol Phys. 1993;26:73–79. 55. Batasakis JG, Rice DH, Solomon AR. The pathology of head and neck tumors. Squamous and mucus gland carcinomas of the nasal cavity, paranasal sinuses, and larynx, part 6. Head Neck Surg. 1980;2:497–508. 56. Ohngren LG. Malignant tumours of the maxillo-ethmoidal region. Acta Otolaryngol Suppl. 1933;19:101–106. 57. Rouviere H. Anatomie Des Lymphatiques De L’homme. Paris: Masson et cie, 1932. 58. Werner JA, Dunne AA, Myers JN. Functional anatomy of the lymphatic drainage system of the upper aerodigestive tract and its role in metastasis of squamous cell carcinoma. Head Neck. 2003;25:322–332. 59. Fernandez JM, Santaolalla F, Del Rey AS, et al. Preliminary study of the lymphatic drainage system of the nose and paranasal sinuses and its role in the detection of sentinel metastatic nodes. Acta Otolaryngol. 2005;125:566–570. 60. Tiwari R, Hardillo JA, Mehta D, et al. Squamous cell carcinoma of maxillary sinus. Head Neck. 2000;22:164–169. 61. Kawashima M, Ogino T, Hayashi R, et al. Influence of postsurgical residual tumor volume on local control in radiotherapy for maxillary sinus cancer. Jap J Clin Oncol. 2001;31:195– 202. 62. Qureshi SS, Chaukar DA, Talole SD, et al. Squamous cell carinoma of the maxillary sinus: A Tata Memorial Hospital experience. Indian J Cancer. 2006;43:26–29. 63. Waldron J, Witterick I. Paranasal sinus cancer: Caveats and controversies. World J Surg. 2003;27:849–855. 64. Le QT, Fu KK, Kaplan MJ, et al. Lymph node metastases in maxillary sinus carcinoma. Int J Radiat Oncol Biol Phys. 2000;46:541– 549.
65. St. Pierre S, Baker SR. Squamous cell carcinoma of the maxillary sinus: Analysis of 66 cases. Head Neck Surg. 1983;5:508–513. 66. Stern SJ, Goepfert H, Clayman G, et al. Squamous cell carcinoma of the maxillary sinus. Arch Otolaryngol Head Neck Surg. 1993;119:964–969. 67. Konno A, Ishikawa K, Terada N, et al. Analysis of long-term results of our combination therapy for squamous cell carcinoma of the maxillary sinus. Acta Otolaryngol Suppl. 1998;537:57–66. 68. Giri SP, Reddy EK, Gemer LS, et al. Management of advanced squamous cell carcinomas of the maxillary sinus. Cancer. 1992;69:657–661. 69. Biswas D, Crank S. Aetiopathology of maxillary swelling— A 3-year prospective study. Eur Arch Otorhinolaryngol. 2007;11:1295–1299. 70. Kraus DH, Lanzieri CF, Wanamaker JR, et al. Complementary use of computed tomography and magnetic resonance imaging in assessing skull base lesions. Laryngoscope. 1992;102;623– 629. 71. Yousem DM. Imaging in sinus disease. In: Kennedy DW, Bolger WE, Zinreich SJ, eds. Diseases of the Sinuses. Diagnosis and Management. London: BC Decker, 2001:129–148. 72. Som PM, Shapiro MD, Biller HF, et al. Sinonasal tumors and inflammatory tissues: Differentiation with MR imaging. Radiology. 1988;167;803–808. 73. Som PM, Dillon WP, Sze G, et al. Benign and malignant sinonasal lesions with intracranial extension: Differentiation with MRI imaging. Radiology. 1989;172;763–766. 74. Lanzieri CF, Shah M, Kraus D, et al. Use of gadoliniumenhanced MR imaging for differentiating mucoceles from neoplasms in the paranasal sinuses. Radiology. 1991;178;425–428. 75. Cox G. Nose and sinuses. In: Wilson JA, ed. Effective Head and Neck Cancer Management. Third Consensus Document, 2002. London: British Association of Otorhinolaryngologists Head and Neck Surgeons, 2002:165–170. 76. Sobin LH, Wittekind C. International Union Against Cancer: TNM Classification of Malignant Tumors. New York, NY: John Wiley & Sons, 2002. 77. Isobe K, Uno T, Hanazawa T, et al. Preoperative chemotherapy and radiation therapy for squamous cell carcinoma of the maxillary sinus. Jpn J Clin Oncol. 2005;35:633–638. 78. Anniko M, Franzen L, Lofroth PO. Long-term survival of patients with paranasal sinus carcinoma. ORL J Otorhinolaryngol Relat Spec. 1990;52:187–193. 79. Jansen EP, Keus RB, Hilgers FJ, et al. Does the combination of radiotherapy and debulking surgery favor survival in paranasal sinus carcinoma. Int J Radiat Oncol Biol Phys. 2000;48:27–35. 80. Blanco AI, Chao KS, Ozyigit G, et al. Carcinoma of the paranasal sinuses: Long-term outcomes with radiotherapy. Int J Radiat Oncol Biol Phys. 2004;59:51–58. 81. Hoppe BS, Stegman LD, Zelefsky MJ, et al. Treatment of nasal cavity and paranasal sinus cancer with modern radiotherapy techniques in the postoperative setting—the MSKCC experience. Int J Radiat Oncol Biol Phys. 2007;67:691–702. 82. Lavertu P, Roberts JK, Kraus DH, et al. Squamous cell carcinoma of the paranasal sinuses: The Cleveland Clinic experience 1977– 1986. Laryngoscope. 1989;99:1130–1136. 83. Nibu K, Sugasawa M, Asai M, et al. Results of multimodality therapy for squamous cell carcinoma of maxillary sinus. Cancer. 2002;94:1476–1482. 84. Nishino H, Miyata M, Morita M, et al. Combined therapy with conservative surgery, radiotherapy, and regional chemotherapy for maxillary sinus carcinoma. Cancer. 2000;89:1925–1932. 85. Lee HB, Hong JP, Kim KT, et al. Orbital floor and infraorbital rim reconstruction after total maxillectomy using a vascularized calvarial bone flap. Plast Reconstr Surg. 1999;104:646–653. 86. Cordeiro PG, Santamaria E. A classification system and algorithm for reconstruction of maxillectomy and midfacial defects. Plast Reconstr Surg. 2000;105:2331–2346. 87. Nakayama B, Hasegawa Y, Hyodo I, et al. Reconstruction using three-dimensional orbitozygomatic skeletal model of titanium mesh plate and soft-tissue free flap transfer following total maxillectomy. Plast Reconstr Surg. 2004;114:631–639.
Chapter 29: Squamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses 88. Ahmad K, Cordoba RB, Fayos JV. Squamous cell carcinoma of the maxillary sinus. Arch Otolaryngol. 1981;107:48–51. 89. Jeremic B, Shibamoto Y, Milicic B, et al. Elective ipsilateral neck irradiation of patients with locally advanced maxillary sinus carcinoma. Cancer. 2000;88:2246–2251. 90. Amendola BE, Eisert D, Hazra TA, et al. Carcinoma of the maxillary antrum: Surgery or radiation therapy? Int J Radiat Oncol Biol Phys. 1981;7:743–746. 91. Lee F, Ogura JH. Maxillary sinus carcinoma. Laryngoscope. 1981;91:133–139. 92. Paulino AC, Marks JE, Bricker P, et al. Results of treatment of patients with maxillary sinus carcinoma. Cancer. 1998;83:457– 465. 93. Guntinas-Lichius O, Kreppel MP, Stuetzer H, et al. Single modality and multimodality treatment of nasal and paranasal sinuses cancer: A single institution experience of 229 patients. Eur J Surg Oncol. 2007;33:222–228. 94. Waldron JN, O’Sullivan B, Warde P, et al. Ethmoid sinus cancer: Twenty-nine cases managed with primary radiation therapy. Int J Radiat Oncol Biol Phys. 1998;41:361–369. 95. Chen AM, Daly ME, Bucci MK, et al. Carcinomas of the paranasal sinuses and nasal cavity treated with radiotherapy at a single institution over five decades: Are we making improvement? Int J Radiat Oncol Biol Phys. 2007;69:141– 147. 96. Karim AB, Kralendonk JH, Njo KH, et al. Ethmoid and upper nasal cavity carcinoma: Treatment, results, and complications. Radiother Oncol. 1990;19:109–120. 97. Parsons JT, Kimsey FC, Mendenhall WM, et al. Radiation therapy for sinus malignancies. Otolaryngol Clin North Am. 1995;28:1259–1268. 98. Parsons JT, Mendenhall WM, Mancuso AA, et al. Malignant tumors of the nasal cavity and ethmoid and sphenoid sinuses. Int J Radiat Oncol Biol Phys. 1988;14:11–22. 99. Katz TS, Mendenhall WM, Morris CG, et al, Malignant tumors of the nasal cavity and paranasal sinuses. Head Neck. 2002;24:821– 829. 100. Dulguerov P, Allal AS. Nasal and paranasal sinus carcinoma: How can we continue to make progress? Curr Opin Otolaryngol Head Neck Surg. 2006;14:67–72. 101. Monroe AT, Bhandare N, Morris CG, et al. Preventing radiation retinopathy with hyperfractionation. Int J Radiat Oncol Biol Phys. 2003;55:776–784. 102. Isaacs JH Jr, Mooney S, Mendenhall WM, et al. Cancer of the maxillary sinus treated with surgery and/or radiation therapy. Am J Surg. 1990;56:327–330. 103. Jiang GL, Ang KK, Peters LJ, et al. Maxillary sinus carcinoma: Natural history and results of postoperative radiotherapy. Radiother Oncol. 1991;21:193–200. 104. Parsons JT, Bova FJ, Fitzgerald CR, et al. Radiation retinopathy after external beam irradiation: Analysis of time-dose factors. Int J Radiat Biol Phys. 1994;30:765–773. 105. Parsons JT, Bova FJ, Fitzgerald CR, et al. Radiation optic neuropathy after megavoltage external beam irradiation: Analysis of time-dose factors. Int J Radiat Biol Phys. 1994;30:755–763. 106. Daly ME, Chen AM, Bucci MK, et al. Intensity-modulated radiation therapy for malignancies of the nasal cavity and paranasal sinuses. Int J Radiat Oncol Biol Phys. 2007;67:151–157. 107. Meyers CA, Geara F, Wong PF, et al. Neurocognitive effects of therapeutic irradiation for skull base tumors. Int J Radiat Oncol Biol Phys. 2000;46:51–55. 108. Duthoy W, Boterberg T, Claus F, et al. Postoperative intensitymodulated radiotherapy in sinonasal carcinoma. Clinical results in 39 patients. Cancer. 2005;104:71–82. 109. Pommier P, Ginester C, Sunyach M, et al. Conformal radiotherapy for paranasal sinus and nasal cavity tumors: Three dimensional treatment planning and preliminary results in 40 patients. Int J Radiat Oncol Biol Phys. 2000;48:485–493. 110. Roa WH, Hazuka MB, Sandler HM, et al. Results of primary and adjuvant CT-based 3-dimensional radiotherapy for malignant tumors of the paranasal sinuses. Int J Radiat Oncol Biol Phys. 1994;28:857–865.
443
111. Weber DC, Chan AW, Lessell S, et al. Visual outcome of accelerated fractionated radiation for advanced sinonasal malignancies employing photons/protons. Radiother Oncol. 2006;81:243– 249. 112. Lomax AJ, Goitein M, Adams J. Intensity modulation of radiotherapy: Photons versus protons in the paranasal sinus. Radiother Oncol. 2003;66:11–18. 113. Hu YH, Tu GY, Qi YQ, et al. Comparison of pre- and postoperative radiation in the combined treatment of carcinoma of the maxillary sinus. Int J Radiat Oncol Biol Phys. 1982;8:1045–1049. 114. Backhouse TW. The treatment of malignant tumors of the maxillary antrum. Clin Radiol. 1986;37:179–182. 115. Curran AJ, Gullane PJ, Waldron J, et al. Surgical salvage after failed radiation for paranasal sinus malignancy. Laryngoscope. 1998;108:1618–1622. 116. Levendag PC, Nijdam WM, von Moolenburgh SE, et al. Interstitial radiation therapy for early stage nasal vestibule cancer: A continuing quest for optimal tumor control and cosmesis. Int J Radiat Oncol Biol Phys. 2006;66:160–169. 117. Mazeron JJ, Chassagne D, Crook J, et al. Radiation therapy of carcinomas of the skin of nose and nasal vestibule: A report of 1676 cases by the Groupe Europeen de Curietherapie. Radiother Oncol. 1988;13:165–173. 118. Sato Y, Morita M, Takahashi H, et al. Combined surgery, radiotherapy, and regional chemotherapy in carcinoma of the paranasal sinuses. Cancer. 1970;25:571–579. 119. Shiga K, Yokoyama J, Hashimoto S, et al. Combined therapy after superselective arterial cisplatin infusion to treat maxillary squamous cell carcinoma. Otolaryngol Head Neck Surg. 2007;136:1003–1009. 120. Samant S, Robbins T, Vang M, et al. Intra-arterial cisplatin and concomitant radiation therapy followed by surgery for advanced paranasal sinus cancer. Arch Otolaryngol Head Neck Surg. 2004;130:948–955. 121. Lee MM, Vokes EE, Rosen A, et al. Multimodality therapy in advanced paranasal sinus carcinoma: Superior long-term results. Cancer J Sci Am. 1999;5:219–223. 122. Rosen A, Vokes EE, Scher N, et al. Locoregionally advance paranasal sinus carcinoma. Favorable survival with multimodality therapy. Arch Otolaryngol Head Neck Surg. 1993;119:743–746. 123. Harrison LB, Pfister DG, Kraus D, et al. Management of unresectable malignant tumors at the skull base using concomitant chemotherapy and radiotherapy with accelerated fractionation. Skull Base Surg. 1994;4:127–131. 124. Itami J, Uno T, Aruga M, et al. Squamous cell carcinoma of the maxillary sinus treated with radiation therapy and conservative surgery. Cancer. 1998;82:104–107. 125. Knegt PP, Ah-See KW, vd Velden LA, et al. Adenocarcinoma of the ethmoidal sinus complex. Surgical debulking and topical fluorouracil may be the optimal treatment. Arch Otolaryngol Head Neck Surg. 2001;127:141–146. 126. Carrau RL, Segas J, Nuss DW, et al. Squamous cell carcinoma of the sinonasal tract invading the orbit. Laryngoscope. 1999;109:230–235. 127. McCary WS, Levine PA, Cantrell RW. Preservation of the eye in the treatment of sinonasal malignant neoplasms with orbital involvement. Arch Otolaryngol Head Neck Surg. 1996;122;657– 659. 128. Perry C, Levine PA, Williamson BR, et al. Preservation of the eye in paranasal sinus cancer surgery. Arch Otolaryngol Head Neck Surg. 1988;114;632–634. 129. Tiwari R, van der Wal J, van der Waal I, et al. Studies of the anatomy and pathology of the orbit in carcinoma of the maxillary sinus and their impact on preservation of the eye in maxillectomy. Head Neck. 1998;20:193–196. 130. Quatela VC, Futran ND, Boynton JR. Eye banking: Techniques for eye preservation in selected neoplasms encroaching on the globe. Otolaryngol Head Neck Surg. 1993;108;662–670. 131. Stern SJ, Goepfert H, Clayman G, et al. Orbital preservation in maxillectomy. Otolaryngol Head Neck Surg. 1993;109:111– 115.
444
Sheahan et al.
132. Imola MJ, Schramm VL. Orbital preservation in surgical management of sinonasal malignancy. Laryngoscope. 2002;112:1357–1365. 133. Andersen PE, Kraus DH, Arbit D, et al. Management of the orbit during anterior fossa craniofacial resection. Arch Otolaryngol Head Neck Surg. 1996;122:1305–1307. 134. Scurry WC, Goldenberg D, Chee MY, et al. Regional recurrence of squamous cell carcinoma of the nasal cavity. A systematic review and meta-analysis. Arch Otolaryngol Head Neck Surg. 2007;133:796–800.
135. Penzer RD, Moss WT, Tong D, et al. Cervical lymph node metastasis in patients with squamous cell carcinoma of the maxillary antrum: The role of elective irradiation of the clinically negative neck. Int J Radiat Biol. 1979;5:1977–1980. 136. Shipchandler TZ, Batra PS, Citardi MJ, et al. Outcomes for endoscopic resection of sinonasal squamous cell carcinoma. Laryngoscope. 2005;115:1983–1987. 137. Buchman L, Larsen C, Pollack A, et al. Endoscopic techniques in resection of anterior skull base/paranasal sinus malignancies. Laryngoscope. 2006;116:1749–1754.
30 Nonsquamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses ` Carlo L. Solero, Stefano Riccio, and Sarah Colombo Giulio Cantu,
In this chapter, we will deal with adenoid cystic carcinoma and, in particular, with adenocarcinoma.
INCIDENCE AND EPIDEMIOLOGY Malignant tumors of the nasal cavity and paranasal sinuses are relatively rare, accounting for 3% of head and neck carcinomas and about 0.5% of all malignancies (1). Despite the low incidence rate, a great variety of histologic types exist. Therefore, the published series from each institution are generally small and heterogeneous, preventing definitive conclusions. Cancers of the paranasal sinuses occur more frequently during the fifth and sixth decades of life, with a male to female ratio of about 2:1 (2). The most common histologic type is squamous cell carcinoma or one of its variants (e.g., transitional, verrucous); adenocarcinoma, adenoid cystic carcinoma, mucoepidermoid carcinoma, and undifferentiated carcinoma are less common. The maxillary sinus is the most frequent site of origin for paranasal sinus tumors (50–70%), particularly squamous cell, adenoid cystic, and mucoepidermoid carcinoma. In the ethmoid sinus, there is a higher incidence rate of undifferentiated carcinoma and adenocarcinoma. The etiology of sinonasal cancers was first hypothesized in 1890, when a tumor was detected in a worker exposed to chrome (3). Since then, many etiologic studies on workers exposed to different materials have been performed. The carcinogenicity of some of these agents in humans has been clearly demonstrated (wood and leather dusts, nickel, chrome, isopropyl alcohol, and arsenic). The role of other work environments in tumor development is more questionable (textile and building industries). For squamous cell carcinoma, smoking is an important risk factor (4). However, the most interesting tumor for which there is an indisputable occupational etiology is intestinal type adenocarcinoma (ITAC); we will present the details in the paragraph dealing with this histologic type.
CLINICAL FEATURES The natural history and clinical features of nonsquamous cell carcinomas obviously depend on the histologic type. The details of each tumor are presented herein. As a rule, the signs and symptoms of nonsquamous cell carcinoma differ from those of squamous cell carcinoma. Many of these are slow-growing tumors, and the first symptoms may date back many months or even years. The history of a submucosal and painless mass may extend over 10 years. Because tumors of the paranasal sinuses arise in air-filled cavities, they can infiltrate the bony walls before signs and symptoms develop. Given this early clinical silence, most patients at presentation have advanced disease extensively involving surrounding structures. We may, indeed, say that malignant tumors of the sinuses do not display clear evidence of their presence until they have broken out of the sinus of origin. Tumors arising in the upper part of the nasal cavity and in the ethmoid may invade the orbit and the anterior cranial fossa through the cribriform plate. They also may destroy the septum and nasal bone, thus infiltrating the skin. The sphenoid sinus and the nasopharynx are invaded in advanced tumors. The possible extension of tumors of the maxillary sinus varies by the site of origin. Lesions arising in the anteroinferior wall often present in the oral cavity as submucosal swelling causing dental pain, loosening of teeth, or improper seating of a denture. Tumors arising on the medial wall may easily invade the nasal cavity through this thin bone. Posterior lesions are the most dangerous for their long clinical silence. Pain is a late-occurring event and often indicates infiltration of the second and/or third branch of trigeminal nerve. The tumor destroys the pterygoid plates and invades the pterygoid and infratemporal fossa. It may approach and erode the greater wing of the sphenoid, spreading to the middle cranial fossa. The incidence of lymph node metastases is low, both at presentation and during follow-up. A higher rate of lymphatic spread occurs only with tumors invading the oral cavity and/or nasal cavity mucosa. This was recognized as far back as 1937 by del Regato (7) and was later confirmed by other authors (8,9). For nonsquamous cell carcinomas, the rate of regional metastases is even lower.
PATHOLOGY Although squamous cell carcinoma is the most common histologic subtype among paranasal sinus malignancies, it is less predominant in this anatomic location than in any other site within the upper aerodigestive tract. Actually, a large variety of histopathologically different tumors occur in this region. The World Health Organization classification divides nasal cavity and paranasal sinus primary malignancies into malignant epithelial tumors, malignant soft tissue tumors, malignant tumors of bone and cartilage, hematolymphoid tumors, and neuroectodermal tumors (5). With regard to epithelial tumors, there are two basic types: those originating from the epithelium and those originating from mucous glands. Nonsquamous cell carcinomas take their origin from the mucous membranes, minor salivary glands, and seromucinous glands (6).
STAGING The difficulties establishing an indisputable prognostic staging for each extension of paranasal sinus carcinomas are demonstrated in the long list of different classifications that existed in times past. These classifications (10–16) considered 445
446
Cantu` et al.
Table 1 T1 T2 T3 T4
INT Staging of Ethmoid Malignant Tumors
Tumor involving the ethmoid and nasal cavity sparing the most superior ethmoid cells Tumor with extension to or erosion of the cribriform plate, with or without erosion of the lamina papyracea and without extension into the orbit Tumor extending into the anterior cranial fossa extradurally and/or into the anterior two-thirds of the orbit, with or without erosion of the anteroinferior wall of the sphenoid sinus, and/or involvement of the maxillary and frontal sinus Tumor with intradural extension, or involving the orbital apex, the sphenoid sinus, the pterygoid plate, the infratemporal fossa, or the skin
only the maxillary sinus, and almost all assigned a higher stage for tumors involving the posterosuperior part of ¨ maxillary sinus. Ohngren line (10) (a plane connecting the inner canthus of the eye to the mandibular angle) divides the maxillary sinus into an anteroinferior portion (infrastructure) and a superoposterior portion (suprastructure). It was used in all versions of the American Joint Committee on Cancer (AJCC) classification of maxillary tumors to distinguish tumors with good prognosis (infrastructure) from tumors with poor prognosis (suprastructure). The International Union Against Cancer (UICC) began to stage maxillary sinus tumors in its fifth edition and used the same criterion. The poorer outcome of superoposterior tumors is the consequence of their early invasion of critical structures such as the orbit, pterygoid, infratemporal fossa, and skull base. Some of these extensions have been considered unresectable for many years, or resectable without wide, clear margins. The introduction of craniofacial resections in routine surgical procedures has enabled skilled surgeons to attain clean margins also in these cases. The AJCC–UICC fifth classification of maxillary sinus carcinomas has been tested in several (including small) series. Results showed the classification to be rather prognostic, with a progressive worsening of the prognosis from T1 to T4. Dulguerov et al. (9), in their meta-analysis of publications on nasal and paranasal sinus carcinoma from 1960 to 2000, found a clear correlation between T stage and survival. Moreover, they demonstrated a nearly unchanged prognosis for T1–T2 tumors during the study period, whereas there was a progressive improvement in outcome for T3–T4 tumors. Even if this result is the logical consequence of the evolution in surgical procedures, one must be aware that the widespread application of modern imaging methods may result in a possible shift in classification from lower to upper stages. This process results in an apparent improvement in outcome by moving the worst cases of a lower stage to the even worse cases of the higher stage, which is the basis of the well-known Will Rogers phenomenon. An example is the division of the stage T4 of the AJCC–UICC-1997 classification (17,18) into stages T4a and T4b of the last 2002 classification (19,20). Thus, the results from upcoming studies should be carefully interpreted, especially when assessing therapeutic improvements. Apart from tumors of the frontal and sphenoid sinuses, where primary tumors are exceptionally rare, the AJCC and UICC had not provided staging guidelines for the more common ethmoid sinus and nasal cavity tumors before 1997. This led to an obvious lack of disease staging in the reported literature on ethmoid cancer. Sisson et al. (21) wrote, “The ethmoid cancers were not staged because there is no generally accepted staging system for this site.” Spiro et al. (22), after having staged tumors of the maxillary sinuses, wrote, “As there is no widely accepted staging system for the remaining sinuses or the nasal cavity, no attempt was made to stage tumors arising in these sites.” Nevertheless, others have attempted to stage nasoethmoid tumors. Kadish et al. (23), Biller et al. (24), and Dulguerov and Calcaterra (25) proposed a classification for
esthesioneuroblastomas. Ellingwood and Million (26) published a classification for cancers of the nasal cavity and ethmoid/sphenoid sinuses in 1979. Finally, Roux et al. (27) adopted a staging system they called “modified TNM.” Moreover, some of the classifications, despite their historical significance as first attempts at staging, were never tested on large series of patients to verify their prognostic value. Ethmoid carcinomas were finally staged in the fifth edition of both the AJCC Cancer Staging Manual (17) and the UICC’s TNM Classification of Malignant Tumours (18). In the absence of a universally accepted staging system and on the basis of our experience with anterior craniofacial resections, we developed in 1993 and presented in 1997 an original classification for malignant ethmoid tumors (28) based on the most commonly accepted unfavorable prognostic factors (involvement of dura mater; intradural extension; involvement of the orbit and, in particular, its apex; invasion of maxillary, frontal, and/or sphenoid sinus; and invasion of the infratemporal fossa and skin) (Table 1). We applied this classification to all consecutive malignant nasoethmoid tumors that were treated at our institution (29). In 1999, we successfully validated our original INT (Istituto Nazionale Tumori) classification for ethmoid cancers (30) against the fifth edition of the AJCC–UICC classification (17,18). On the basis of these encouraging results, we tested the sixth AJCC–UICC (19,20) 2002 classification in terms of prognostic performance versus the 1997 AJCC–UICC and the INT classifications (31). Both the 1997 and 2002 AJCC–UICC classifications seemed to have limited prognostic value. By contrast, the INT classification satisfied one of the main goals of tumor staging, demonstrating the progressive worsening of prognosis with different tumor classes. The validity of the INT classification was confirmed on our large series of 241 patients, and it was shown to achieve the best prognostic discrimination among T classifications not only for the overall series but also when applied separately to untreated patients, recurring cases, and adenocarcinomas, the most frequent histologic type in our series. Dulguerov et al., in their recent review (32), stated, “While the evolution of TNM staging is a work in continuous progress, the T staging of ethmoid and nasal primaries needs an urgent revision.” We agree with this statement. After providing these general comments on paranasal sinuses malignancies, we now discuss two histologic types among nonsquamous cell carcinomas.
ADENOID CYSTIC CARCINOMA Incidence and Epidemiology Adenoid cystic carcinoma is the most frequent malignant tumor of the minor salivary glands, constituting more than one-third of cases. According Harrison and Lund (33), it represents about 1.5% of all tumors of the paranasal sinuses; the maxilla and hard palate are the most frequent sites of origin. In our series of 704 cases of malignant tumors of the paranasal sinuses, we found 115 cases of adenoid cystic carcinoma (16.4%): 24/305 cases (7.9%) in the ethmoid sinus and
Chapter 30: Nonsquamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses
91/399 (22.8%) in the maxillary sinus. In a series of 334 patients who underwent a craniofacial resection collected from 17 institutions, Ganly et al. (34) reported 32 cases of minor salivary glands carcinoma (9.6%). There is no known etiologic factor for this tumor. The male-to-female ratio is different in each published series and is likely equal. In our series, there was a small prevalence of females (65/50). The age range is quite broad, but there are peaks in the fifth and sixth decades of life.
Pathology Adenoid cystic carcinoma has no distinguishing gross features apart from its infiltrative growth, which makes it difficult to demarcate the tumor from surrounding normal tissues (6). Three histologic types have been described: tubular, cribriform, and solid. Batsakis et al. (35,36) suggested that the solid type is the most aggressive form but other authors have not found an indisputable correlation between histology and outcome (37,38). Perineural spread is the distinctive feature of adenoid cystic carcinoma; it may extend great distance from the primary tumor along nerve pathways.
Clinical Features The clinical features of adenoid cystic carcinoma depend on the site of origin. Because minor salivary glands are rare in the anterior part of the hard palate, the tumor often appears as a submucosal mass involving the posterior hard palate and/or soft palate. Many of these lesions are indolent and painless until ulceration appears, so the history may go back many months or even years. Tumors that present in the maxillary sinus may be associated with facial swelling and pain. Pain, tingling, or paresthesias indicate neural involvement. Bearing in mind the aforesaid capacity for perineural spread, it is easy to understand how extension of maxillary tumors along the second trigeminal branch (greater palatine and infraorbital nerves) can reach intracranial spaces, in particular the Gasserrian ganglion and cavernous sinus (Fig. 1). Ethmoid tumors can easily involve the anterior cranial fossa along olfactory nerves. Traversing the dura, the tumor may extend into the brain. In case of orbital invasion, in particular of the apex, the tumor may follow the first branch of trigeminal nerve; the third, fourth, and sixth cranial nerves; and the optic nerve.
447
Lymph node metastases are rare, and lymphatic spread plays a modest role in the treatment of adenoid cystic carcinoma. Only one patient in our series died from nodal metastases per se. On the contrary, systemic metastases (lung, brain, bone, and liver) are frequent. Many patients manifest metastatic disease concurrent with or after local recurrence within 5 years, and about 5% of patients have lung metastases at presentation. However, distant metastases without local recurrence may appear 15 to 20 years later (33).
Treatment Surgery and postoperative radiotherapy is usually considered the treatment of choice (39). Only small palatal lesions can be resected with partial or subtotal maxillectomy. Unfortunately, the majority of patients present with large maxillary tumors are either approaching or involving the infratemporal fossa, the orbit, and the middle and/or anterior skull base. Obviously, the goal of surgery must be complete resection with negative margins. Technical advances in anterior, lateral, and anterolateral craniofacial resections with free flap reconstruction allow the resection of tumors that were considered inoperable in the past. However, oncologic resection of paranasal sinuses adenoid cystic carcinoma often has an uncertain outcome for the aforesaid factors. The surgeon must be aware that even aggressive surgery typically does not result in cure. Local relapse and/or distant metastases are nearly always the rule. Thus, a balance between radical surgery and low morbidity must be sought. In our opinion, the patient’s clinical condition before treatment must be the main factor in making a decision. Acute pain from trigeminal infiltration, or the presence of an ulcerated, necrotic, and bleeding tumor, may justify a wide resection and reconstruction to give to the patient a better quality of life, sometimes for many years. The role of radiotherapy alone in the treatment of sinonasal adenoid cystic carcinoma is controversial. Current opinion is that this tumor is not radiocurable with conventional megavoltage photon and/or electron beams (33). Some reports indicate that neutron radiotherapy might be more efficacious than conventional radiotherapy (40,41). However, these articles underline unsatisfactory outcomes in patients with skull base involvement, as this extension is dose limiting because of the sensitivity of the central nervous system structures to neutron radiotherapy. Chemotherapy is generally reserved for palliative treatment of metastatic disease or locoregional recurrence for which further surgery or radiation is not possible (42).
Outcome and Prognosis
Figure 1 Maxillary adenoid cystic carcinoma involving the third branch of trigeminal nerve.
Because of the long natural history of this tumor, a 5-year follow-up period is inadequate to convey the ultimate outcome of adenoid cystic carcinoma. Patients may have frequent local recurrences and/or hematogenous dissemination, sometimes 10 or more years later. Because few papers report a large number of patients with a long follow-up, it is difficult to assess the real outcome for patients presenting with adenoid cystic carcinoma of the sinonasal tract involving the skull base. Despite aggressive surgery and postoperative radiotherapy, about 70% of patients will experience a tumor recurrence, and the cure rate is even lower for patients treated for local recurrence after previous surgery (43). In our series of 115 patients with adenoid cystic carcinoma of the paranasal sinuses, 52 patients presented with a T4a-b tumor involving the skull base. While 7 of 24 patients who underwent an anterior craniofacial resection for a tumor localized in the ethmoid sinus are free
448
Cantu` et al.
of disease (29%), none of the 28 patients requiring a lateral craniofacial resection for involvement of the middle cranial fossa is alive without disease. We believe that although surgery appears to be palliative for many patients with advanced adenoid cystic carcinoma, a tumor that is bleeding and causing pain may justify a wide craniofacial resection and reconstruction to give to the patient a better quality of residual life.
ADENOCARCINOMA Pathology, Incidence, and Epidemiology The true incidence of sinonasal adenocarcinomas is unknown because of the lack of consensus among pathologists regarding its classification. While almost all pathologists clearly distinguish between salivary types and true sinonasal adenocarcinomas, most of published clinical articles on paranasal sinus tumors use the general term adenocarcinoma without subclassification (9,44–46). Batsakis (47) categorized adenocarcinomas into three clinicopathologic forms: papillary, sessile, and alveolarmucoid, the last of which includes tumors that closely simulate colonic carcinomas. Abecasis et al. (48) used different clinicopathologic and immunohistochemical classifications: high- and low-grade adenocarcinoma, papillary, and ITAC. In that classification, ITACs were given a poor prognosis. Heffner et al. (49) divided adenocarcinomas of the sinonasal tract into low grade and high grade, and included ITACs in the high-grade group. In contrast, Bashir et al. (50) classified ITACs as well- or moderately differentiated tumors. The last paper is paradigmatic about the difficulties of histologic classification. These authors classified 11 sinonasal adenocarcinomas into three groups: intestinal type, glandular type, and solid type. However, after this statement, the authors wrote: “Judging from other articles, some of our tumors that we have classified as SNA-G (glandular) and/or SNA-S (solid) may be considered intestinal type by others.” Barnes (51) stated that ITACs of the nasal cavity and paranasal sinuses might occur sporadically or as an occupational disease, in cases of wood dust exposure. Histologically, he recognized five variants: papillary, colonic, solid, mucinous, and mixed. Comparing 17 cases of sporadic-type ITAC to those among woodworkers, he found some important differences. In the former group, there were nine men and eight women, with eight tumors originating in the maxillary sinus, seven in the nasal cavity, and only two in the ethmoid sinus. In contrast, ITACs in woodworkers occurred primarily in men and originated almost exclusively in the nasal cavity or ethmoid sinus. Immunohistochemical marking and expression of oncoproteins did not completely provide resolution. Bashir et al. (50) wrote, “The study of expression of CK7 and CK20 in sinonasal adenocarcinoma is not useful in making the accurate differential diagnosis between primary or metastatic intestinal tumors. The CK7/CK20 profile was successful in distinguishing adenocarcinoma from transitional carcinoma, pointing to its utility in the differential diagnosis of these two entities.” We may find similar conclusions in the paper by Abecasis et al. (48). According to Choi et al. (52), “All primary enteric-type carcinomas and the 2 colonic metastases were reactive to CK20, but all non-enteric-type tumors were negative for CK20 and positive for CK7.” These authors conclude, “Non-enteric-type (seromucinous) adenocarcinoma may originate directly from surface respiratory-type epithelium or from seromucous glands, metaplastic transformation of surface respiratory to enteric-type epithelium pre-
cedes the development of enteric adenocarcinoma, and coordinate analyses of CK7 and CK20 reactivity may aid the differential diagnosis of adenocarcinoma in the sinonasal tract.” Kennedy et al. (53) stained 12 sinonasal adenocarcinomas with monoclonal antibodies to CK7, CK20, CDX-2, and villin. The authors’ conclusion was, “Sinonasal ITACs have a distinctive phenotype, with all cases expressing CK20, CDX-2, and villin. Most ITACs also express CK7, although a proportion of tumors are CK7 negative. ITAC seems to be preceded by intestinal metaplasia of the respiratory mucosa, which is accompanied by a switch to an intestinal phenotype.” As a consequence of these difficulties in pathologic classification, the literature contains a wide range of incidence rates for adenocarcinomas among sinonasal malignant tumors [e.g., 4–9% for Harrison and Lund (33), 10–20% for Bashir (50)]. With regard to epidemiology and etiology, almost all authors report the role of wood dust. This role was recognized in 1970 by Hadfield (54), who analyzed 35 cases of sinonasal adenocarcinoma in woodworkers in the furniture industries. In all 35 patients, the tumor appeared to originate in the ethmoid sinuses. Since then, several papers have been published on this matter, and the leather and shoe industries have also been associated with an increased risk of adenocarcinoma (55–57). The different roles that hardwoods and softwoods play in the development of these tumors present a vexing question. Some authors (58,59) in northern Europe (where furniture industries use softwoods) underlined a minor and different carcinogenic quality of softwoods compared with hardwoods that are more often used in southern Europe. General consensus does exist, however, that wood dust levels above 5 mg/m3 in the work environment present higher risk. Interestingly, the rates of adenocarcinoma in series of patients with sinonasal tumors treated with anterior craniofacial resection vary greatly between Europe and North America. The rates of adenocarcinoma in European series are very high: Roux (60) (France), 74%; Suarez (61) (Spain), 53%; Cantu (31) (Italy), 49%; and Cheesman (62) (United Kingdom), 27%. In contrast, the rates in American series are much lower: McCutcheon (63) (United States), 17%; Bentz (64) (United States), 12%; Donald (65) (United States), 6%; and Irish (66) (Canada), 5%. Bridger (67) (Australia) reports a rate of 37% of adenocarcinoma, similar to the rates in Europe. It is difficult to find an unambiguous and exhaustive explanation for these discrepancies. We may advance some hypotheses: The high-risk threshold commonly accepted for wood dust level in the air is 5 mg/m3 ; it is likely that this threshold was exceeded in many European artisan furniture factories and joineries in the past. Hardwoods, which present higher risk than softwoods, are probably more widespread in Europe than in America. Safety measures such as the use of masks and aspiration devices became widespread in the United States before Europe. Considering that the latency period from the beginning of exposure to clinical evidence of tumor is about 40 years (68) and that factory conditions have improved in Europe, we may predict a possible reduction in the incidence of the disease in Europe in coming decades. We investigated the exposure to wood dust or leather dust in 499 patients with sinonasal tumors treated at the National Cancer Institute of Milan between 1987 and 2001 (69). Of 249 patients with ethmoid tumors, 124 had adenocarcinomas (115 males and 9 females), and 107 of these patients (86.3%) had previously been exposed to wood or
Chapter 30: Nonsquamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses
leather dusts. Only 2 of the 125 patients with ethmoid tumors other than adenocarcinomas had been exposed to these dusts. Adenocarcinomas were 17 among 250 nonethmoid sinonasal tumors; no exposure to wood or leather dusts was reported in any of these patients. The first remarkable outcome of our study was the confirmation, on a larger series, of Hadfield’s remark; (54) the ethmoid sinus is the only paranasal site where a relationship between adenocarcinoma and wood or leather exposure was evident. Actually, in many epidemiologic papers, we may find general terms like “sinonasal adenocarcinoma,” “nasal cancers,” or sinonasal cancers.” Our study confirms a very week correlation between wood or leather dust exposure and maxillary sinus tumors or ethmoid neoplasms other than ITAC. The second noteworthy result was the duration of exposure. Among males, 87 of 104 patients had been exposed to wood or leather dust for many years during their work life (25–55 years); on the contrary, 17 patients had been exposed for just a short period (less than 10 years), and many years before the clinical evidence of the tumor (23–47 years). This result was unexpected because the analysis of 12 casecontrol studies demonstrated that the risk for ethmoid adenocarcinomas is proportional to the duration of exposure to the oncogenic agents (70). By contrast, our data suggest that even a short period of exposure followed by a long latency may be sufficient for gene deregulation, which leads to the onset of the disease. Previous work from our institution (71) demonstrated a high percentage of TP53, p14ARF and p16aINK4 deregulation, and H-ras mutations in patients with ethmoid adenocarcinoma exposed to wood or leather dusts, thus supporting the epidemiologic observation of a genotoxic origin of this tumor.
Clinical Features Sinonasal adenocarcinoma may present differently based on the site of origin. In the rare cases originating in the maxillary sinus, the symptoms are similar to those of other histotypes. However, because most adenocarcinomas originate in the ethmoid sinus, the symptoms are often nonspecific and innocuous: unilateral nasal obstruction, rhinorrhea, and epistaxis (49,51,56). Pain, epiphora, and other orbital symptoms are less common and often occur late in development of the tumor. Because these symptoms frequently occur in many sinonasal diseases, ITACs are often not detected until the tumor is very large. Most patients in published series had a T3–T4 tumor. Rarely the tumor may present with a glabellar mass involving the frontal bones. The first symptom reported by our 167 patients with ethmoid adenocarcinoma who underwent an anterior craniofacial resection was in decreasing order of occurrence: nasal obstruction, 94 patients (56%); epistaxis, 48 patients (29%); and rhinorrhea, 10 patients (6%). Pain, anosmia, exophthalmos, glabellar swelling, and visual disturbances were rare. Anosmia is a very interesting symptom; even if only 3 patients reported it as the first symptom that led the patient to seek medical care, most patients remembered a partial or total anosmia some years before the beginning of nasal obstruction or epistaxis. As reported by other authors (49,56), duration of symptoms in our patients ranged from 1 to 30 months. ITAC presents as an exophytic pink mass bulging into the nasal cavity, often with a gray, necrotic, and friable appearance. Coronal CT and/or MRI with contrast enhancement are mandatory to define the extent of the disease, in particular for orbital and intracranial involvement.
449
Cervical node metastases are rare (33). Only two patients in our series had nodal involvement at presentation (1.2%). Five patients developed node metastases during follow-up, together with a relapse of the primary; however, none of them died from cervical metastases per se. The situation is different for maxillary sinus. Le et al. (72) report a 25% rate of nodal involvement for maxillary sinus adenocarcinomas, in agreement with our results (22%). Probably in some series of maxillary sinus tumors there are also nonintestinal type cases, and these tumors act like adenocarcinomas of salivary glands. Distant metastases are infrequent (33,56). Only three patients in our series had developed lung metastases (two of them together with a relapse of the primary).
Treatment Because of the rarity of ethmoid adenocarcinoma, it is nearly impossible to compare the various treatment options in a clinical trial. However, surgical radical resection is the most frequent primary treatment for patients with this tumor (9,32,73). Because most of these tumors approach or involve the cribriform plate, anterior craniofacial resection is the established “gold standard” (34,45,46,60–67,74). Some authors have demonstrated an improvement in disease-free survival with craniofacial resection compared with transfacial resection alone (33,45,74). Interest in endoscopic surgery for malignant tumors of the anterior skull base is increasing. Although most articles on this surgical approach involve esthesioneuroblastoma (75–77), some authors advocate completely endoscopic resection for ethmoid adenocarcinomas also (78,79). We believe that the resection of the sinonasal component of the tumor, however it is done, must be radical. Because ITAC is often an occupational disease with likely wide field mucosal changes, at least a total ethmoidectomy must be performed. The metaplastic transformation of ethmoid mucosa to enteric-type epithelium precedes the development of enteric adenocarcinoma (52,53), and possible preneoplastic foci may be present in macroscopically uninvolved sites of ethmoid. Moreover, ITAC is a locally aggressive tumor that easily infiltrates the underlying bone (47). As the main characteristic of endoscopic approach is the subperiosteal resection (77), we wonder if such resection may be suitable for this tumor. We have already seen three cases of ITAC arising in unresected parts of the ethmoid after an endoscopic resection performed elsewhere (Fig. 2). Other authors stress a combined endoscopic and intracranial approach to avoid any external facial incision (80,81). Because of the infrequency of nodal involvement, a prophylactic neck dissection in patients with N0 disease is not indicated (82). It is difficult to establish the results of radiotherapy alone in the treatment of patients with sinonasal adenocarcinoma. Most published series probably have some selection bias, because favorable lesions are treated with surgery, leaving larger tumors to radiotherapy. However, except as seen in a few publications, the results of radiation alone are poorer than those with treatments that include surgery (32). On the contrary, there is a general consensus about the use of radiotherapy in combination with surgery (9,32,73). The most effective sequence for surgery and radiotherapy has not been definitively determined. Although most authors prefer primary surgery (29,32,44,61,63,67), some (83) continue to choose primary radiotherapy with surgery for salvage. However, these authors report a very high rate of visual complications, with 20 of 29 patients (69%) developing blindness or impaired vision in at least one eye.
450
Cantu` et al.
resection; and postoperative radiotherapy. Twelve patients achieved a pCR, and 18 did not (overall response rate, 40%). In patients with wild-type TP53 or functional p53 protein, the pCR rates were 83% and 80%, respectively; in patients with mutated TP53 or impaired p53 protein, the pCR rates were 11% and 0%, respectively. At a median 55-month follow-up, all pCR patients were disease free; 44% of nonresponding patients had a relapse. These results are indicative of the probable existence of two genetic ITAC subgroups. The differences in TP53 mutation status or protein functionality strongly influence pCR after chemotherapy and prognosis. In spite of these encouraging results, we must remember that only 30% to 40% of ITAC patients proved to be chemoresponders; moreover, because the goal of this study was possible avoidance of surgery in chemoresponders, in our experience it was impossible to predict pCRs before surgery, even with the use of CT scan, MRI, and PET imaging. Knegt et al. (88) proposed a very unusual treatment protocol: surgical debulking via an extended anterior maxillary antrostomy followed by a combination of repeated topical chemotherapy (5-fluorouracil emulsion) and necrotomy. However, their good results have never been reproduced elsewhere. Figure 2 Relapse of ethmoid adenocarcinoma after endoscopic resection. The tumor involves unresected parts of the ethmoid.
The role of chemotherapy (intravenous or intraarterial) remains unclear. Many reports do not distinguish paranasal sinus carcinomas by specific subsites and histology. In cases in which a distinction has been made, there are few cases of adenocarcinoma (84). The only published article involving a number of ethmoid sinus adenocarcinomas suggested that chemotherapy associated with craniofacial resection and radiotherapy could improve the overall treatment outcome, and patients achieving a good clinical, and especially a pathologic, response appeared to obtain the greatest benefit (85). Based on these indications from the literature, a prospective phase II study in 49 patients with paranasal sinus cancer (47/49 with ITAC) was conducted in our institution to investigate the role of primary chemotherapy within the multidisciplinary approach to these tumors (86). After 3 to 5 cycles of chemotherapy (leucovorin, 5-fluorouracil, and cisplatin), 42 patients underwent an anterior craniofacial resection and postoperative radiotherapy. All gross specimens were carefully evaluated with at least 20 to 25 tumor sections. A complete pathologic remission (pCR) was found in eight cases (16%), and all patients achieving pCR were free of disease at a median follow-up of 26 months. In light of these results, we tried to find biologic markers able to predict the response to primary chemotherapy, to maximize treatment benefit and to avoid unnecessary toxicity in patients treated with potentially ineffective drugs (87). Because it was demonstrated (71) that the presence of TP53 mutations is one of the main genetic hallmarks of ITACs, considering the correlation between ITAC and professional exposure to wood or leather dusts, the link between genotoxic exposition and loss of p53 function, and the relationship between TP53 functional status and response to DNA-damaging treatment, we investigated 30 patients with ITAC, assessing the TP53 gene mutation profile on pretreatment biopsy. All patients underwent primary chemotherapy with cisplatin, 5-fluorouracil, and leucovorin; craniofacial
Outcome and Prognosis It is not easy to report indisputable data on outcome and prognosis for patients with ethmoid adenocarcinoma. Unfortunately, many published reports describing treatment and outcome do not distinguish paranasal sinus carcinomas in specific subsites and histology. We may find ethmoid cancers with nasal cavity and maxillary sinus tumors (44, 45,48, 49). Those few papers dealing with ethmoid tumors often contain a small number of cases and/or different histologies (46,50,52,53,56,83). Heffner et al. (49) classified their cases as low- and high-grade adenocarcinoma, reporting a good prognosis for the former and a poor outcome for the later. Barnes (51) stated 40% overall survival for patients with professional ITAC. Choi et al. (52) noted a higher local recurrence in patients with the enteric-type adenocarcinoma. In the series of Abecasis et al. (48), intestinal-type tumors were associated with a worse prognosis than were transitional tumors. Overall, in precraniofacial resection period, the reported outcomes were disappointing (33,56). Because the most important prognostic factor is local extension, the introduction of anterior craniofacial resection improved the cure rates (33,45,74). Considering that the lack of a widespread clinical prognostic classification does not allow a clear comparison among published series, we may find very different cure rates: Bridger (67), 70%; Bentz (64), 68%; Orvidas (45) and Howard (74), 58%; and Suarez (61), 31%. Ganly et al. (34) report a 5-year overall survival of 44.8% for patients with adenocarcinoma of paranasal sinuses, and who are treated with anterior craniofacial resection in 17 different institutions. However, we must remember that 5-year survival figures may fail to give the real picture of ITACs, because some patients develop a recurrence after more than 5 years (33). In Howard’s series (74), the 58% 5-year overall survival dropped to 40% at 10 years. Also Roux et al. (60) report an overall survival of 51% and 23% at 5 and 10 years, respectively. Knegt’s results (88) are astonishing (disease-free survival of 87% at 5 years and 74% at 10 years). Few papers have classified patients with ITAC in clinical stages. Roux (60) used a modified TNM staging; overall 10-year survival rates for his patients were 75% for T3, 38% for T4a, and 0% for T4b.
Chapter 30: Nonsquamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses
Five-year overall survival rate for our patients was 43.7%. Categorizing the cases according to the INT staging system (31), the corresponding rates were as follows: 61% for T2, 49% for T3, and 10% for T4. Stage by stage, we found lower cure rates for patients who underwent craniofacial resection after previous surgery performed elsewhere, in comparison with untreated patients. In conclusion, sinonasal adenocarcinomas encompass a variety of histologic types. It is difficult to declare an indisputable association between tumor histology and patient survival. Instead, survival is clearly related to intracranial extension. Anterior craniofacial resection and postoperative radiotherapy remain the standard treatment, and previous inadequate resection may jeopardize the results. The use of endoscopic resection remains controversial. As relapses may occur after many years, a long-term follow-up is crucial for patient’s survival.
REFERENCES 1. Muir CS, Nectoux J. Descriptive epidemiology of malignant neoplasms of nose, nasal cavities, middle ear and accessory sinuses. Clin Otolaryngol Allied Sci. 1980;5(3):195–211. 2. Rousc GC. Epidemiology of cancer of the nose and paranasal sinuses: Current concepts. Head Neck Surg. 1979;2(1):3–11. 3. Newman DA. A case of adeno-carcinoma of the left inferior turbinated body, and perforation of the nasal septum, in the person of a worker in chrome pigments. Glasgow Med J. 1890;33:469– 470. 4. Mannetje A, Kogevinas M, Luce D, et al. Sinonasal cancer, occupation, and tobacco smoking in European women and men. Am J Ind Med. 1999;36(1):101–107. 5. Barnes L, Eveson J, Reichart P, et al. World Health Organization Classification of Tumours: Pathology and Genetics–Head and Neck Tumours. Lyon, France: IARC Press, 2005. 6. Batsakis JG, Rice DH, Solomon AR. The pathology of head and neck tumors. Squamous and mucous gland carcinomas of the nasal cavity, paranasal sinuses and larynx. Part 7. Head Neck Surg. 1980;2(6):497–508. 7. del Regato JA. Roentgentherapy in epitheliomas of the maxillary sinus. Surg Gynecol Obstet. 1937;65:657–665. 8. Kim GE, Chung EJ, Lim JJ, et al. Clinical significance of neck node metastasis in squamous cell carcinoma of the maxillary antrum. Am J Otolaryngol. 1999;20(6):383–390. 9. Dulguerov P, Jacobsen MS, Allal AS, et al. Nasal and paranasal sinus carcinoma: Are we making progress? Cancer. 2001;92(12):3012–3129. ¨ 10. Ohngren LG. Malignant tumors of the maxillo-ethmoidal region: A clinical study with special reference to the treatment with electrosurgery and irradiation. Acta Otolaryngol (Stockolm). 1933;19:1–476. 11. Sisson GA, Johnson NE, Amiri CS. Cancer of the maxillary sinus. Clinical classification and management. Ann Otol Rhinol Laryngol. 1963;72(12):1050–1059. 12. Sakai S, Hamasaki Y. Proposal for the classification of carcinoma of the paranasal sinuses. Acta Otolaryngol (Stockolm). 1967;63(1):42–48. 13. Lederman M. Cancer of the upper jaw and nasal chambers. Proc R Soc Med. 1969; 62(1):65–72. 14. Rubin P. Cancer of the head and neck: Nose, paranasal sinuses. JAMA. 1972;219(3):336–338. 15. Chandler JR, Guillamondegui OM, Sisson GA, et al. Clinical staging of cancer of the head and neck: A new “new” system. Am J Surg. 1976;132(4):525–528. 16. Harrison DF. Critical look at the classification of maxillary sinus carcinomata. Ann Otol Rhinol Laryngol. 1978;87(1 Pt 1):3–9. 17. American Joint Committee on Cancer. AJCC Cancer Staging Manual, 5th ed. Philadelphia, PA: Lippincot-Raven, 1997. 18. Sobin LH, Wettekind C, eds. TNM Classification of Malignant Tumors, 5th ed. New York, NY: John Wiley & Sons Inc, 1997.
451
19. American Joint Committee on Cancer. AJCC Cancer Staging Manual, 6th ed. New York, NY: Springer-Verlag, 2002. 20. UICC. TNM Classification of Malignant Tumours, 6th ed. New York, NY: Wiley-Liss, 2002. 21. Sisson GA, Toriumi DM, Atiyah RA. Paranasal sinus malignancy: A comprehensive update. Laryngoscope. 1989;99(2):143– 150. 22. Spiro JD, Soo KC, Spiro RH. Nonsquamous cell malignant neoplasms of the nasal cavities and paranasal sinuses. Head Neck. 1995;17(2):114–118. 23. Kadish S, Goodman M, Wang CC. Olfactory neuroblastoma. A clinical analysis of 17 cases. Cancer. 1976;37(3):1571–1576. 24. Biller HF, Lawson W, Sachdev VP, et al. Esthesioneuroblastoma: Surgical treatment without radiation. Laryngoscope. 1990;100(11):1199–1201. 25. Dulguerov P, Calcaterra T. Esthesioneuroblastoma: The UCLA experience 1970–1990. Laryngoscope. 1992;102(8):843–849. 26. Ellingwood KE, Million RR. Cancer of the nasal cavity and ethmoid/sphenoid sinuses. Cancer. 1979;43(4):1517–1526. 27. Roux FX, Brasnu D, Menard M, et al. Les abords combin´es des tumeurs malignes de l’ethmo¨ıde et autres sinus paranasaux. Principes et r´esultats. Ann Otolaryngol Chir Cervicofac. 1991;108(5):292–297. 28. Cantu G, Solero CL, Salvatori P, et al. A new classification of malignant ethmoid tumors. 3rd European Skull Base Congress. London, 9–11 April 1997. Skull Base Surg. 1997;7(Supp2): 33. 29. Cantu` G, Solero CL, Mariani L, et al. Anterior craniofacial resection for malignant ethmoid tumors–A series of 91 patients. Head Neck. 1999;21(3):185–191. 30. Cantu G, Solero CL, Mariani L, et al. A new classification for malignant tumors involving the anterior skull base. Arch Otolaryngol Head Neck Surg. 1999;125(11):1252–1257. 31. Cantu G, Solero CL, Miceli R, et al. Which classification for ethmoid malignant tumors involving the anterior skull base? Head Neck. 2005;27(3):224–231. 32. Dulguerov P, Allal AS. Nasal and paranasal sinus carcinoma: How can we continue to make progress? Curr Opin Otolaryngol Head Neck Surg. 2006;14(2):67–72. 33. Harrison D, Lund VJ. Non epidermoid epithelial neoplasms. In: Harrison D, Lund VJ, eds. Tumors of the Upper Jaw. Edinburgh, UK: Churchill Livingstone, 1993:105–134. 34. Ganly I, Patel SG, Sing B, et al. Craniofacial resection for malignant paranasal sinus tumors: Report of an international collaborative study. Head Neck. 2005;27(7):575–584. 35. Batsakis JG, Luna MA, el-Naggar A. Histopathologic grading of salivary gland neoplasms: III. Adenoid cystic carcinomas. Ann Otol Rhinol Laryngol. 1990;99(12):1007–1009. 36. Matsuba HM, Spector GJ, Thawley SE, et al. Adenoid cystic salivary gland carcinoma. A histopathologic review of treatment failure patterns. Cancer. 1986;57(3):519–524. 37. Spiro RH, Huvos AG, Strong EW. Adenoid cystic carcinoma of salivary origin. A clinicopathologic study of 242 cases. Am J Surg. 1974;128(4):512–520. 38. Pitman KT, Prokopakis EP, Aydogan B, et al. The role of skull base surgery for the treatment of adenoid cystic carcinoma of the sinonasal tract. Head Neck. 1999;21(5):402–407. 39. Mendenhall WM, Morris CG, Amdur RJ, et al. Radiotherapy alone or combined with surgery for adenoid cystic carcinoma of the head and neck. Head Neck. 2004;26(2):154–162. 40. Douglas JG, Laramore GA, Austin-Seymour M, et al. Treatment of locally advanced adenoid cystic carcinoma of the head and neck with neutron radiotherapy. Int J Radiat Oncol Biol Phys. 2000;46(3):551–557. 41. Douglas JG, Koh W, Ausin-Seymour M, et al. Treatment of salivary gland neoplasms with fast neutron radiotherapy. Arch Otolaryngol Head Neck Surg. 2003;129(9):944–948. 42. Laurie SA, Licitra L. Systemic therapy in the palliative management of advanced salivary gland cancers. J Clin Oncol. 2006;24(17):2673–2678. 43. Pitman KT, Prokopakis EP, Aydogan B, et al. The role of skull base surgery for the treatment of adenoid cystic carcinoma of the sinonasal tract. Head Neck. 1999;21(5):402–407.
452
Cantu` et al.
44. Katz TS, Mendenhall WM, Morris CG, et al. Malignant tumors of the nasal cavity and paranasal sinuses. Head Neck. 2002;24(9):821–829. 45. Orvidas LJ, Lewis JE, Weaver AL, et al. Adenocarcinoma of the nose and paranasal sinuses: A retrospective study of diagnosis, histologic characteristics, and outcomes in 24 patients. Head Neck. 2005;27(5):370–375. 46. Wax MK, Yun KJ, Wetmore SJ, et al. Adenocarcinoma of the ethmoid sinus. Head Neck. 1995;17(4):303–311. 47. Batsakis JG. Mucous gland tumors of the nose and paranasal sinuses. Ann Otol Rhinol Laryngol. 1970;79(3):557–562. 48. Abecasis J, Viana AG, Pissarra C, et al. Adenocarcinomas of the nasal cavity and paranasal sinuses: A clinicopathological and immunohistochemical study of 14 cases. Histopathology. 2004;45(3):254–259. 49. Heffner DK, Hyams VJ, Hauck KW, et al. Low-grade adenocarcinoma of the nasal cavity and paranasal sinuses. Cancer. 1982;50(2):312–322. 50. Bashir AA, Robinson RA, Benda JA, et al. Sinonasal adenocarcinoma: immunohistochemical marking and expression of oncoproteins. Head Neck. 2003;25(9):763–771. 51. Barnes L. Intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Am J Surg Pathol. 1986;10(3):192–202. 52. Choi HR, Sturgis EM, Rashid A, et al. Sinonasal adenocarcinoma: evidence for histogenetic divergence of the enteric and nonenteric phenotypes. Hum Pathol. 2003;34(11):1101–1107. 53. Kennedy MT, Jordan RCK, Berean KW, et al. Expression pattern of CK7, CK20, CDX-2, and villin in intestinal-type sinonasal adenocarcinoma. J Clin Pathol. 2004;57(9):932–937. 54. Hadfield EH. A study of adenocarcinoma of the paranasal sinuses in woodworkers in the furniture industry. Ann R Coll Surg Engl. 1970;46(6):301–319. 55. Acheson ED. Nasal cancer in the furniture and boot and shoe manufacturing industries. Prev Med. 1976;5(2):295–315. 56. Klintenberg C, Olofsson J, Hellquist H, et al. Adenocarcinoma of the ethmoid sinuses. A review of 28 cases with special reference to wood dust exposure. Cancer. 1984;54(3):482–488. 57. Luce D, Leclerc A, Morcet JF, et al. Occupational risk factors for sinonasal cancer: A case-control study in France. Am J Ind Med. 1992;21(2):163–175. 58. Hernberg S, Westerholm P, Shultz-Larsen K, et al. Nasal and sinonasal cancer. Connection with occupational exposures in Denmark, Finland and Sweden. Scand J Work Environ Health. 1983;9(4):315–326. 59. Voss R, Stenersen T, Roald Hoppedal B, et al. Sinonasal cancer and exposure to softwood. Acta Otolaryngol. 1985;99(1–2):172– 188. 60. Roux FX, Pages JC, Nataf F, et al. Malignant ethmoidsphenoidal tumors. 130 cases. Retrospective study. Neurochirurgie. 1997;43(2):100–110. 61. Suarez C, Llorente JL, Fernandez DeLeon R, et al. Prognostic factors in sinonasal tumors involving the anterior skull base. Head Neck. 2004;26(2):136–144. 62. Cheesman AD, Craniofacial resection for tumors of the nasal cavity and paranasal sinuses. Head Neck Surg. 1986;8(6):429– 435. 63. McCutcheon IE, Blacklock JB, Weber RS, et al. Anterior transcranial (craniofacial) resection of tumors of the paranasal sinuses: surgical technique and results. Neurosurgery. 1996;38(3):471–479; discussion 479–480. 64. Bentz BG, Bilsky MH, Shah JP, et al. Anterior skull base surgery for malignant tumors: A multivariate analysis of 27 years of experience. Head Neck. 2003;25(7):515–520. 65. Donald PJ. Complications in skull base surgery for malignancy. Laryngoscope. 1999;109(12):1959–1966. 66. Irish JC, Gullane PJ, Gentili F, et al. Tumors of the skull base: Outcome and survival analysis of 77 cases. Head Neck. 1994;16(1):3– 10.
67. Bridger GP, Kvok B, Baldwin M, et al. Craniofacial resection for paranasal sinus cancers. Head Neck. 2000;22(8):772–780. 68. Nylander LA, Dement JD. Carcinogenetics effects of wood dust: Review and discussion. Am J Ind Med. 1993;24(5):619–647. 69. Bimbi G, Saraceno MS, Riccio S, et al. Adenocarcinoma of ethmoid sinus: An occupational disease. Acta Otorhinolaryngol Ital. 2004;24(4):199–203. 70. Demers PA, Kogevinas M, Boffetta P, et al. Wood dust and sinonasal cancer: pooled reanalysis of twelve case-control studies. Am J Ind Med. 1995;28(2):151–166. 71. Perrone F, Oggionni M, Birindelli S, et al. TP53, p14ARF, p16INK4a and H-ras gene molecular analysis in intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Int J Cancer. 2003;105(2):196–203. 72. Le QT, Fu KK, Kaplan MJ, et al. Lymph node metastasis in maxillary sinus carcinoma. Int J Radiat Oncol Biol Phys. 2000;46(3):541– 549. 73. Waldron J, Witterick I. Paranasal sinus cancer: Caveats and controversies. World J Surg. 2003;27(7):849–855. 74. Howard DJ, Lund VJ, Wei WI. Craniofacial resection for tumors of the nasal cavity and paranasal sinuses: A 25-year experience. Head Neck. 2006;28(10):867–873. 75. Casiano RR, Numa WA, Falquez AM. Endoscopic resection of esthesioneuroblastoma. Am J Rhinol. 2001;15(4):271–279. 76. Unger F, Haselsberger K, Walch C, et al. Combined endoscopic surgery and radiosurgery as treatment modality for olfactory neuroblastoma (esthesioneuroblastoma). Acta Neurochir (Wien). 2005;147(6):595–601. 77. Castelnuovo PG, Delu G, Sberze F, et al. Esthesioneuroblastoma: Endonasal endoscopic treatment. Skull Base. 2006;16(1):25–30. 78. Haberman W, Zanarotti U, Groell R, et al. Combination of surgery and gamma knife radiosurgery–a therapeutic option for patients with tumors of nasal cavity or paranasal sinuses infiltrating the skull base. Acta Otorhinolaryngol Ital. 2002;22(2):74–79. 79. Bockmuhl U, Minovi A, Kratzsch B, et al. Endonasal microendoscopic tumor surgery: State of the art. Laryngorhinootologie. 2005;84(12):884–891. 80. Thaler ER, Kotapka M, Lanza DC, et al. Endoscopically assisted anterior cranial skull base resection of sinonasal tumors. Am J Rhinol. 1999;13(4):303–310. 81. Castelnuovo PG, Belli E, Bignami M, et al. Endoscopic nasal and anterior craniotomy resection for malignant nasoethmoid tumors involving the anterior skull base. Skull Base. 2006;16(1):15–18. 82. Jiang GL, Morrison WH, Garden AS, et al. Ethmoid sinus carcinomas: Natural history and treatment results. Radiother Oncol. 1998;49(1):21–27. 83. Waldron JN, O’Sullivan B, Warde P, et al. Ethmoid sinus cancer: Twenty-nine cases managed with primary radiation therapy. Int J Radiat Oncol Biol Phys. 1998;41(2):361–369. 84. Samant S, Robbins KT, Vang M, et al. Intra-arterial cisplatin and concomitant radiation therapy followed by surgery for advanced paranasal sinus cancer. Arch Otolaryngol Head Neck Surg. 2004;130(8):948–955. 85. Brasnu D, Laccourreye O, Bassot V, et al. Cisplatin-based neoadjuvant chemotherapy and combined resection for ethmoid sinus adenocarcinoma reaching and/or invading the skull base. Arch Otolaryngol Head Neck Surg. 1996;122(7):765–768. 86. Licitra L, Locati LD, Cavina R, et al. Primary chemotherapy followed by anterior craniofacial resection and radiotherapy for paranasal cancer. Ann Oncol. 2003;14(3):367–372. 87. Licitra L, Suardi S, Bossi P, et al. Prediction of TP53 status for primary cisplatin, fluorouracil, and leucovorin chemotherapy in ethmoid sinus intestinal-type adenocarcinoma. J Clin Oncol. 2004;22(24):4901–4906. 88. Knegt PP, Ah-See KW, vd Velden LA, et al. Adenocarcinoma of the ethmoidal sinus complex: Surgical debulking and topical fluorouracil may be the optimal treatment. Arch Otolaryngol Head Neck Surg. 2001;127(2):141–146.
31 Esthesioneuroblastoma Valerie J. Lund and David J. Howard
preponderance is more pronounced with a 2:1 ratio and the age range is 12 to 70 years (mean 46 years), though children under 10 have occasionally been reported (16)(Fig. 1).
INCIDENCE AND EPIDEMIOLOGY In common with all tumors of the anterior skull base, esthesioneuroblastoma (ENB) is comparatively rare. This malignant neuroendocrine neoplasm arose from the olfactory mucosa and was first recognized by Berger et al. in 1924 (1), who coined the term “esthesioneuroepitheliome olfactif.” However, a wide range of other terms has been used including esthesioneurocytoma, esthesioneuroma, intranasal neuroblastoma, olfactory neuroepithelial tumor, and olfactory neuroblastoma. In 1966, Skolnik et al. (2) found only 97 cases reported in 42 papers in the English literature with most authors only treating 2 or 3 cases, and by 1989 O’Connor estimated that ≤300 cases had been published which represented 1% to 5% of all malignant tumors of the nasal cavity (3). However, this number had risen to 945 by 1997 (4), a report which did not include a large series from the Armed Forces Institute of Pathology (5) nor from the Institut Gustave-Roussy (6). In 2000, the National Cancer Data Base included 664 cases from over 500 U.S. hospitals over a 10-year period (1985–1995). In recent years, increasing numbers of this tumor are being described, almost certainly due to an increasing awareness and improved histological techniques for diagnosis. Despite this, it remains difficult to accrue large individual series, compromising statistical analysis of outcome. The authors, working in a tertiary referral center, have had the opportunity of managing 78 cases since 1970. Hitherto, unlike adenocarcinoma, occupational factors in the development of olfactory neuroblastoma have not been identified in men other than a single case report in a woodworker (7). However, in rodents the administration of N-nitroso compounds has been reported to produce esthesioneuroepitheliomas (8–10) when administered parenterally, orally, or topically, as has administration of bischloromethyl ether (11). To evaluate possible etiological factors, the occupational history and exposure to possible carcinogens was examined in a series of 54 patients with olfactory neuroblastoma using questionnaires and/or structured interviews. This revealed four individuals (8 %) who were dental practitioners (2) or dental nurses (2), whereas no other members of the dental profession have been found in a cohort of over 700 other sinonasal malignancies (12). It is neither clear what chemical, if any, is implicated nor does this constitute sufficient evidence to determine a definite link but is worthy of continued observation. No hereditary patterns have been described in this tumor and there is no apparent racial predilection. In the literature, there appears to be a slight male preponderance and the tumor may occur over a wide age range (3–90 years) with a reported bimodal peak in the second/third decades and sixth/seventh decades (3,13–15). In our own series, this male
PATHOLOGY Olfactory neuroblastoma generally arises in the nasal roof, corresponding to the anatomical distribution of the olfactory epithelium, which extends from the olfactory niche onto the upper nasal septum and superior turbinates on the lateral wall. Evidence for the origin of olfactory neuroblastoma from specialized olfactory epithelium, however, is somewhat circumstantial (5), though tumors occasionally found outside this distribution have been ascribed to ectopic olfactory epithelium. Tumors arising in the cribriform niche can easily spread superiorly along olfactory fibers into the anterior cranial fossa to affect the olfactory bulb and tracts. The superior septum is often involved and from thence the tumor may spread to the contralateral side and into the ethmoids and adjacent orbit. Histological studies suggest that there is microscopic intracranial involvement in the majority of patients even when not suggested by imaging and the macroscopic appearances at surgery (17). Macroscopically, the tumor is characteristically a polypoid reddish gray mass that bleeds readily. Microscopically, the tumor typically forms clusters of cells arranged in patterns, which vary from small nests surrounded by a fibrillary stroma to diffuse areas separated by fibrovascular septa. The cells may palisade around blood vessels and occasionally true rosettes form. It had been suggested that ENB was part of the Ewing sarcoma/peripheral neuroectodermal group of tumors but this has not been supported by immunohistochemical studies (18). However, it continues to present some difficulties in diagnosis even with modern techniques and can be confused with a host of other small-cell tumors such as lymphoma, malignant melanoma, and undifferentiated sinonasal carcinoma by those unfamiliar with sinonasal malignancy. This prompted Ogura and Schenck to describe ENB as the “great imposter” (19). Immunohistochemistry using a broad panel of antibodies is usually employed to confirm the diagnosis. These include general neuroendocrine markers such as neuronspecific enolase, synaptophysin, chromogranin, and protein gene product–9.5, which are usually positive (20). S-100 positivity can be demonstrated at the periphery of the tumor nests and some tumors are also positive using MNF 116 and CAM 5.2, both stains for certain cytokeratins. Conversely LP 34, a high molecular weight cytokeratin stain, epithelial membrane antigen, carcinoembryonic antigen, and glial fibrillary acidic protein are generally negative. 453
454
Lund and Howard
The age distribution of the olfactory tumors 8 7
Number of cases
6 5 4 3 2 1 0
11–15 16–20 21–25 26–30 31–35 36–40 41–45 46–50 51–55 56–60 61–65 66–70 71–76 76–80 Age (years)
Figure 1 Histogram showing distribution of patients by age in authors’ cohort.
Other features such as microfilaments, microtubules, and neurosecretory granules may be seen on electron microscopy. Indeed, this technique was used in the days before immunohistochemistry to confirm diagnosis. Histological features have been used by some authors to determine prognosis, with limited success (5,21,22).
STAGING A number of staging systems have been proposed. Kadish et al. (23) is a somewhat crude system dividing the tumors into three stages. Stage A: lesions confined to the nasal cavity Stage B: involvement of nasal cavity + one or more of the paranasal sinuses Stage C: involvement beyond the nasal cavity including the orbit, skull base, intracranial cavity, cervical lymph nodes, or systemic metastases. This can now be considerably refined by the use of modern imaging protocols (24) validated by craniofacial resection. However, neither the Kadish system nor subsequent modifications (5,22,25) have proved entirely successful, largely due to the late presentation of most patients, despite efforts to refine advanced disease by creating a Stage D for metastases.
MANAGEMENT Clinical Features The usual site of origin results in fairly innocuous symptoms initially, only remarkable by their sudden onset and unilaterality. As a consequence, there is often considerable delay in diagnosis, with some patients waiting for more than a year [24% of 40 cases reported by Schwabb et al. (6)]. There is little specific to this particular tumor, the symptoms being common to all nasal cavity lesions, that is, blockage, discharge, some epistaxis due to the tumor’s vascularity, and hyposmia. In a series of 42 patients, unilateral obstruction occurred in 93%, epistaxis in 55%, and rhinorrhea in 30% (15). Anosmia was reported rather rarely (5%) in our patients and invasion of the anterior cranial fossa is otherwise generally silent. As the tumor spreads to the orbit, patients may get epiphora, displacement of the eye, diplopia, and eventually visual loss,
though this is usually a late phenomenon. Ocular symptoms occurred in 11% of our patients (15). Curiously, in this series the left side was more often affected (62%) compared to the right (29%), with both sides affected at presentation in 9%. Occasionally, involvement of the Eustachian orifice may result in otalgia and a conductive hearing loss. The incidence of cervical metastases varies considerably from report to report, compounded by the generally small numbers in each series (14). A retrospective review (26) suggested an incidence of 27% based on 10 series, which included 207 cases. However, if only Kadish Stage C was considered, the rate increased to 44% but subsequent review suggests that some of the data in this study was incorrect. In 1993, Harrison and Lund (27) found only one case out of 20 in contrast to Morita et al. (25), who reported 20.4% (10/49) in a series from the Mayo Clinic. Reviewing 320 cases, reported in 15 series, Rinaldo et al. (14) reported a lymph node metastatic frequency of 23.4% (range 5–100%), though only eight studies validated the diagnosis of ENB with immunohistochemistry. As ENB is a neuroendocrine tumor, it can be associated with syndromes caused by inappropriate hormone production such as Cushing or antidiuretic secretion (6,28–31).
Imaging All patients are ideally submitted to a preoperative imaging protocol, which employs a combination of high-resolution contrast-enhanced computer tomography (coronal and axial planes) combined with multiplanar magnetic resonance imaging (MRI) enhanced with gadolinium diethylenetriamine pentaacetic acid (DTPA) (24). Following surgery, all patients should also undergo a rigorous follow-up protocol of MRI, ideally combined with sinonasal endoscopy (with biopsy of any suspicious lesions), every 3 to 4 months for the first 2 years and then 6 monthly thereafter (32). Patients may develop recurrence many years after initial treatment and this can be anywhere in relation to the surgical field, orbit, or intracranial cavity suggesting some local embolic phenomenon. Appropriate imaging using a combination of computed tomography (CT) and magnetic resonance imaging (MRI) will often suggest the diagnosis but more importantly indicate extent (24,33). There are no features which are specific to ENB, though the position of the mass and associated bone erosion indicates a malignant nasal tumor (Fig. 2). Coronal CT remains the most accurate method of demonstrating early anterior skull base erosion, while the addition of contrast enhancement and MRI shows extent of intracranial and orbital spread. Typical features are an intense signal on precontrast T2-weighted spin echo sequences and strong enhancement after gadolinium on T1-weighted sequences. However, even the most sophisticated imaging cannot be absolutely relied upon to demonstrate involvement of the dura and orbital periosteum, which can only be determined by surgery with histological confirmation. Ultrasound of the neck combined with fine-needle aspiration cytology is an important screening technique with a high degree of accuracy (34). Although metastatic spread can occur outside the head and neck, it is rare at presentation, and, therefore, it is not the authors’ practice to undertake more extensive body imaging at this stage.
Treatment The advent of craniofacial resection has revolutionized the treatment of ENB, doubling survival figures and is now regarded as the “gold standard.” In a meta-analysis, Dulgerov et al. confirmed that this procedure combined with radiotherapy was the treatment of choice (13).
Chapter 31: Esthesioneuroblastoma
(A)
455
(C)
(B)
(D)
Figure 2 Imaging of olfactory neuroblastoma. (A) Coronal CT showing small tumor confined to nasal cavity. (B) Coronal MRI in same patient (T1 with gadolinium) confirming extent of lesion and suitable for endoscopic resection. (C) Coronal CT showing larger tumor apparently confined to nasal cavity in another patient. (D) Contemporaneous coronal MRI in same patient (T1 with gadolinium) showing obvious extension into anterior cranial fossa requiring craniofacial resection. (C) Coronal CT showing larger tumor apparently confined to nasal cavity in another patient. (D) Contemporaneous coronal MRI in same patient (T1 with gadolinium) showing obvious extension into anterior cranial fossa requiring craniofacial resection.
Surgery Craniofacial resection was introduced in the 1970s by Ketcham and others, providing the combination of an en bloc oncologic resection with low morbidity and excellent cosmesis (35–37). By approaching the tumor from the nose and anterior cranial fossa, the operation directly addresses the origin and local spread of this tumor, allowing resection of dura and the olfactory system including the olfactory epithelium, cribriform plate, olfactory bulb, and tracts. This directly deals with macro- and microscopic spread of disease, reducing local recurrence (38,39). There are many variations on the technique but essentially all involve some form of craniotomy, together with a nasal approach using various incisions and forms of repair.
Using a coronal incision in the scalp and a sublabial incision for a midfacial degloving, these scars can be hidden, though the use of an extended lateral rhinotomy or a supraorbital spectacle incision heals well. The skull base repair may be affected with a pericranial flap or fascia lata and split skin. In our series which extended over 27 years, using this technique of craniofacial resection in 308 patients, postoperative hospital stay has been on average 14 days and major complications low (39) (Table 1). Prior to craniofacial resection, conventional wisdom dictated that the orbit should be sacrificed if tumor was either adjacent or had transgressed the periosteum. However, it is clear that it is possible to salvage a proportion of these eyes without compromising survival. If the tumor has not
456
Lund and Howard Table 1 Complications Associated with Craniofaciala and Endoscopic Resection Complication Immediate DVT CSF leak Long term Glaucoma Epilepsy Epiphora Serous otitis media Frontal bone necrosis None Total cohort
Craniofacial n
Endoscopic n
Chemotherapy
1 1 1 1 1 1 3 37 42
advent of intensity modulation radiotherapy may of some advantage in this respect. Generally, radiotherapy has been used as an adjunct to surgery (15,49) and postoperative delivery is preferred.
1
10 11
a Complications multiple in some cases. Abbreviations: DVT, deep-vein thrombosis; CSF, cerebrospinal fluid.
penetrated through the full thickness of orbital periosteum on frozen section, it is possible to resect it widely and skin graft the area. Nonetheless, the orbit should be cleared if there is full thickness periosteal penetration or frank infiltration of orbital contents. More recently, endoscopic resection has been used for selected cases, usually for those who are without significant skull base erosion or for patients who are a poor anaesthetic risk. The endoscopic approach should not be regarded as a limited procedure, rather as the nasal component of the craniofacial resection (40) that includes a wide-field resection of tissue. The main difference is that this is undertaken in a more piecemeal fashion but is performed under excellent direct visualization. The ability to perform skull base resection and repair via an endoscopic approach has facilitated this (41), and, indeed, some surgeons have advocated undertaking extensive intracranial resection from below in teams which involve neurosurgical input (42). Alternatively, the endoscopic approach can be combined with an external craniotomy (43). Patients should be apprised that both might be required or that an endoscopic approach may need to be extended to include a formal craniotomy. In a recent series of 49 patients undergoing endoscopic resection for malignant sinonasal tumors, 11 were olfactory neuroblastomas but follow-up was a mean of only 36 months (40). While all are presently alive, one has been converted to craniofacial resection at 1 year. In other publications, cohorts of malignant tumors including olfactory neuroblastoma have been described (44–48) but, as in this series, follow-up was relatively short. However, complication rates, if any, are very low, hospital stay on average 5 days and postoperative radiotherapy may be started very promptly (Table 1). It should also be noted that endoscopic surgery could have a role following conventional craniofacial resection in the management of localized recurrence. The neck is not traditionally treated prophylactically, although a selective neck dissection is undertaken in the presence of disease.
Radiotherapy Radiation in this area must be carefully administered to deliver the maximum dose while preserving the adjacent brain and optic nerves. An external megavoltage beam and threefield technique has generally been used. An anterior port combined with wedged lateral fields delivers a dose of 55 to 65 Gy. Because of the proximity of the optic chiasm, the dose given must remain below normal tissue tolerance and the
The use of concomitant chemotherapy has not been fully evaluated, though chemosensitivity has been found in retrospective series. ENB has been shown to respond to platinumbased regimes (50–54). In 8 of 18 of our patients between 1999 and 2005 who received radiotherapy and cisplatin, results suggested a reduced recurrence rate, though the number of patients precludes any reliable statistical conclusions (55).
OUTCOME AND PROGNOSIS Prior to craniofacial resection, the use of lateral rhinotomy and radiotherapy provided poor results (56–58) of ≤40% at 5 years, largely due to the inability to deal with intracranial spread. Craniofacial resection specifically addresses this area and allows removal of the olfactory bulbs and tracts where microscopic disease may be residing undetected. Thus when large series with long-term follow-up after craniofacial are considered, the 5-year survival is seen to have improved significantly (59–63) or even doubled as in our own series to 77% (15) or to 89% in that of Diaz et al. (59). However, there is continued loss over time and local recurrence can occur up to 12 years after treatment (range 12–144 months, mean 37 months). In our series of 42 cases, disease-free survival drops from 77% at 5 years to 53% at 10 years. In a further study of this cohort enlarged to 56 individuals, 15-year survival fell to 40%, which emphasizes the importance of long-term follow-up. The most frequent recurrence is local and occurred in 17% of our series in keeping with other published series using craniofacial resection and radiotherapy. Local recurrence has been shown to be decreased by the addition of radiotherapy, but it does not seem to matter whether this is given before or after surgery, even when the therapeutic interval differs between pre- and postoperative administration. This applies to both survival (p = 0.515) and complications (p = 0.07). Interestingly, previous treatment did not seem to affect 5-year actuarial survival either, but in the patients who developed local recurrence, 5-year survival after further salvage treatment was 54%. However, it should again be noted that the site of “local” recurrence can be anywhere and on either side of the nose, sinuses, orbit, or intracranial cavity, so follow-up must be especially vigilant if this disease is to be detected early. Multivariate analysis of the craniofacial series shows that involvement of the brain and orbit are independent factors affecting outcome (15). When survival was considered according to orbital involvement, 5-year actuarial survival was 97% when the eye was not affected and 49% when the periosteum was affected, but when the eye was frankly infiltrated even when the eye was sacrificed, there were no 5-year survivors (p = 0.0067) (Fig. 3). When skull base and intracranial involvement were considered, there was a difference between those whose tumor was confined to the nasal cavity, when compared with those in whom the skull base was affected; those in whom the olfactory tracts were involved; those in whom the dura was additionally infiltrated; and those in whom the brain was affected (p = 0.035). When patients with tumor in the
Chapter 31: Esthesioneuroblastoma
457
cial resection specifically addresses the site of origin and local spread of the tumor and when combined with radiotherapy offers the optimum treatment, which has doubled survival as compared with previous treatments. However, endoscopic resection may offer an alternative in selected cases, though comparable numbers and follow-up are lacking as yet. Thus, the results of endoscopic surgery must be compared with the large craniofacial series with 10 or more years of follow-up as the natural history of this tumor extends over a lifetime. Whichever surgical approach is used, when the orbital periosteum is affected without penetration into the orbital structures themselves, the eye can be preserved without an adverse effect on survival.
REFERENCES
Figure 3 Kaplan Meier survival for orbital involvement in olfactory neuroblastoma. Source: From Ref. 15.
nasal cavity and/or skull base were compared with the other groups, as might be expected there was a statistical difference between these patients and those with dura and/or olfactory tract involvement (p = 0.006) and those with brain involvement (p = 0.039) (Fig. 4). Those patients treated by endoscopic approaches are too limited in number and follow-up to make any meaningful comparison at present, but it is clear that in carefully selected individuals, early results are at least as good as craniofacial resection (40,44–47) and the complication rate and morbidity commensurately less (64). However, given the case mix, this is to be expected. Cervical lymphadenopathy constitutes an important prognostic factor. Koka et al. (31) showed a 29% survival with nodes versus 64% without nodes, and this was supported by two subsequent meta-analyses (13,14). Distant metastases with locoregional control are relatively rare [≤10% (15,54)] and carry a poor prognosis.
CONCLUSION ENB is a rare nasal tumor with a unique natural history. Its management requires experience in histopathology, radiology, sinonasal surgery, and medical oncology. Craniofa-
Figure 4 Kaplan Meier survival for skull base and intracranial involvement in olfactory neuroblastoma. Source: From Ref. 15.
1. Berger L, Luc R, Richard D. L’estheioneuroepitheliome olfactif. Bull Assoc Fr Etude Cancer. 1924;13:410–421. 2. Skolnik EM, Massari FS, Tenta LT. Olfactory neuroepithelioma. Arch Otolaryngol. 1966;84:644–653. 3. O’Connor TA, McLean P, Juillard GS. Olfactory neuroblastoma. Cancer. 1989;63:2426–2428. 4. Broich G, Pagliaria, Ottaviani F. Esthesioneuroblastoma: A general review of the cases published since the discovery of the tumour in 1924. Anticancer Res. 1927;17:2683–2706. 5. Hyams VJ. Olfactory neuroblastoma. In: Hyams VJ, Baksakis JG, Michaels L, eds. Tumours of the Upper Respiratory Tract and Ear. Washington, DC: Armed Forces Institute of Pathology, 1998:240– 248. 6. Schwaab G, Michaeu C, Le Guillou C, et al. Olfactory esthesioneuroma: A report of 40 cases. Laryngoscope. 1988;98:872–876. 7. Magnavita N, Sacco A, Bevilacqua L, et al. Aesthesioneuroblastoma in a woodworker. Occup Med. 2003;53:231–234. 8. Magee PN, Montesano R, Preussmann R. N-nitroso compounds and related carcinogens. In: Searle CE, ed. Chemical carcinogens. Washington, DC: USA American Chemical Society, 1976:491–625. 9. Herrold K. Induction of olfactory neuroepithelia tumours in Syrian hamsters by diethlynitrosamine. Cancer. 1964;17:114–121. 10. Vollrath M, Altmannsberger M, Weber K, et al. Chemically induced tumours of rat olfactory epithelium: A model for human esthesioneuroepithelioma. J Natl Cancer Inst. 1986;76:1205–1169. 11. Leong BK, Kociba RJ, Jersey GC. A lifetime study of rats and mice exposed to vapour of bis(chloromethyl)ether. Toxicol Appl Pharmacol. 1981;58(2):269–281. 12. Personal communication with VJ Lund. 13. Dulguerov P, Abdelkarim SA, Calcaterra TC. Esthesioneuroblastoma: A meta-analysis and review. Lancet Oncol. 2001;2:683–688. 14. Rinaldo A, Ferlito A, Shaha AR, et al. Esthesioneuroblastoma and cervical lymph node metastases: Clinical and therapeutic implications. Acta Otolaryngol. 2002;122:125–221. 15. Lund VJ, Howard D, Wei W, et al. Olfactory neuroblastoma. Laryngoscope. 2003;113:502–507. 16. Kumar M, Fallon R, Hill J, et al. Esthesioneuroblastoma in children. J Pediatr Hematol Oncol. 2002;24:482–487. 17. Harrison D. Surgical pathology of olfactory neuroblastoma. Arch Otolaryngol. 1984;7:60–64. 18. Argani P, Perez-Ordonez B, Xiao H, et al. Olfactory neuroblastoma is not related to the Ewing family of tumors: Absence of EWS/FL11 gene fusion and MIC2 expression. Am J Surg Pathol. 1998;22:391–398. 19. Ogura J, Schenck N. Unusual nasal tumors. Problems in diagnosis and treatment. Otolaryngol Clin N Am. 1973;6:813–837. 20. Lund VJ, Milroy C. Olfactory neuroblastoma: Clinical and pathological aspects. Rhinology. 1993;31:1–6. 21. Hirose T, Scheithauer BW, Lopes MB, et al. Olfactory neuroblastoma: An immunohistochemical, ultrastructural, and flow cytometric study. Cancer. 1995;76:4–19. 22. Papadaki H, Kounelis S, Kapadia S, et al. Relationship of p53 gene alterations with tumor progression and recurrence in olfactory neuroblastoma. Am J Surg Pathol. 1996;20:715–721.
458
Lund and Howard
23. Kadish S, Goodman M, Wang C. Olfactory neuroblastoma: A clinical analysis of 17 cases. Cancer. 1976;37:1571–1576. 24. Lloyd GAS, Lund VJ, Howard DJ, et al. Optimum imaging for sinonasal malignancy. J Laryngol Otology. 2000;114;557–562. 25. Morita A, Ebersold M, Olsen K, et al. Esthesioneuroblastoma: Prognosis and management. Neurosurgery. 1993;32:706–715. 26. Davis R, Weissler M. Esthesioneuroblastoma and neck metastasis. Head Neck. 1992;14:477–482. 27. Harrison DNF, Lund VJ. Neuroectodermal lesions. In: Tumours of the Upper Jaw, London, Edinburgh, UK: Churchill Livingstone. 1993:295–328. 28. Singh W, Ramage C, Best P, et al. Nasal neuroblastoma secreting vasopressin. Cancer. 1980;45:961–966. 29. Srigley J, Dayal V, Gregor R, et al. Hyponatremia secondary to olfactory neuroblastoma. Arch Otolaryngol. 1983;109:559– 562. 30. Myers S, Hardy D, Wiebe C, et al. Olfcatory neuroblastoma invading the oral cavity in a patient with inappropriate antidiuretic hormone secretion. Oral Surg Oral Med Oral Pathol. 1994;77:645– 650. 31. Koka V, Julieron M, Bourhis J, et al. Aesthesioneuroblastoma. J Laryngol Otol. 1998;112:628–633. 32. Lund VJ, Howard DJ, Wei WI, et al. Craniofacial resection for tumors of the nasal cavity and paranasal sinuses—a 17-year experience. Head Neck. 1998;20;97–105. 33. Pickuth D, Heywang-Kobrunner S, Spielman R. Computed tomography and magnetic resonance imaging features of olfactory neuroblastoma: An analysis of 22 cases. Clin Otolaryngol. 1999;24;457–461. 34. Collins B, Cramer H, Hearn S. Fine needle aspiration cytology of metastatic neuroblastoma. Acta Cytol. 1997;41:802–810. 35. Ketcham AS, Van Buren JM. Tumors of the paranasal sinuses: A therapeutic challenge. Am J Surg. 1985;150:406–413. 36. Clifford P. Transcranial approach to cancer of the antro-ethmoidal area. Clin Otolaryngol. 1977;2:115–130. 37. Terz JJ, Young HF, Lawrence W Jr. Combined craniofacial resection for locally advanced carcinoma of the head and neck. Am J Surg. 1980;140:613–624. 38. Shah JP, Kraus DH, Bilsky MH, et al. Craniofacial resection for malignant tumors involving the anterior skull base. Arch Otolaryngol Head Neck Surg. 1997;123:1312–1317. 39. Howard DJ, Lund VJ, Wei WI. Craniofacial resection for sinonasal neoplasia—a twenty-five year experience. Head Neck. 2006;28:867–873. 40. Lund VJ. Endoscopic resection of malignant tumours of the nose and sinuses. Am J Rhinol Fast Track. 2007;21:89–94. 41. Lund VJ. Endoscopic management of CSF leaks. Am J Rhinol. 2002;16:17–23. 42. Kassam A, Horwitz M, Welch W, et al. The role of endoscopic assisted microneurosurgery (image fusion technology) in the performance of neurosurgical procedures. Minim Invasive Neurosurg. 2005;48:191–196. 43. Thaler ER, Kotapka M, Lanza D, et al. Endoscopically assisted anterior cranial skull base resection of sinonasal tumors. Am J Rhinol. 1999;13:303–310. 44. Draf W, Schick B, Weber R, et al. Endoscopic micro-endoscopic surgery of nasal and paranasal sinus tumours. In: Stamm AC,
45.
46.
47. 48.
49.
50.
51.
52. 53.
54.
55.
56. 57.
58. 59. 60. 61. 62.
63.
64.
Draf W, eds. Micro-Endoscopic Surgery of the Paranasal Sinuses and the Skull Base. Berlin, Germany: Springer, 2000:481–488. Goffart Y, Jorissen M, Daele J, et al. Minimally invasive endoscopic management of malignant sinonasal tumours. Acta Otorhinolaryngol Belg. 2000;54:221–232. Walch C, Stammberger H, Anderhuber W, et al. The minimally invasive approach to olfactory neuroblastoma: Combined endoscopic and stereotactic treatment. Laryngoscope. 2000;110;635– 640. Casiano RR, Numa WA, Falquez AM. Endoscopic resection of esthesioneuroblastoma. Am J Rhinol. 2001;15;271–279. Batra PS, Citardi MJ, Worley S, et al. Resection of anterior skull base tumors: Comparison of combined traditional an endoscopic techniques. Am J Rhinol. 2005;19;521–528. Foote RL, Morita A, Ebersold M, et al. Esthesioneuroblastoma: The role of adjuvant radiation therapy. Int J Radiat Oncol Biol Phy. 1993;27:835–842. McElroy E, Buckner J, Lewis J. Chemotherapy for advanced esthesioneuroblastoma: The Mayo Clinic experience. Neurosurgery. 1998;42:1023–1028. Kim D, Jo Y, Heo D, et al. Neoadjuvant etoposide, ifosafamide and cisplatin for the treatment of olfactory neuroblastoma. Cancer. 2004;101:2257–2260. Dulguerov P, Calcaterra T. Esthesioneuroblastoma: The UCLA experience 1970–1990. Laryngoscope. 1992;102:843–849. Eden BV, Debo RF, Larner JM, et al. Esthesioneuroblastoma: Long-term outcome and patterns of failure—the University of Virginia experience. Cancer. 1994;73:2556–2562. Levine PA, Gallagher R, Cantrell RW. Esthesioneuroblastoma: Reflections of a 21-year experience. Laryngoscope. 1999;109:1539– 1543. Nikapota A, Sevitt T, Lund VJ, et al. Outcomes of radical conformal radiotherapy and concomitant cisplatin chemotherapy for olfactory neuroblastoma–a review of a single centre experience. Abstr Am Soc Clin Oncol. 2006. Shah J, Feghali J. Esthesioneuroblastoma. Am J Surg. 1981;142:456–458. Appelblatt N, McClatchey K. Olfactory neuroblastoma: A retrospective clinicopathological study. Head Neck Surg. 1982;5:108– 113. Bailey B, Barton S. Olfactory neuroblastoma. Arch Otolaryngol. 1975;101:1–5. Eriksen J, Bastholt L, Krogdahl A, et al. Esthesioneuroblastoma. What is the optimal treatment? Acta Oncol. 2000;39:231–235. Resto V, Eisele D, Forastiere A, et al. Esthesioneuroblastoma: The Johns Hopkins experience. Head Neck. 2000;55:550–558. Patel SG, Singh B, Polluri A, et al. Craniofacial surgery for malignant skull base tumors. Cancer. 2003;98:1179–1187. Constantinidis J, Steinhart H, Koch M, et al. Olfactory neuroblastoma: The University of Erlangen-Nuremberg experience 1975– 2000. Otolaryngol Head Neck Surg. 2004;130:567–574. Diaz E, Johnigan R, Pero C, et al. Olfactory neuroblastoma. The 22 year experience at one comprehensive cancer center. Head Neck. 2005;27:138–149. Ganly I, Patel SG, Singh B, et al. Complications of craniofacial resection for malignant tumours of the skull base: Report of an international collaborative study. Head Neck. 2005;27;445–451.
32 Melanoma of the Nasal Cavity and Paranasal Sinuses Ziv Gil, Mark H. Bilsky, and Dennis H. Kraus
that among 84,836 cases of malignant melanoma only 1074 cases involved the mucosal membrane, half of them arising in the head and neck (1). According to a report by the Armed Forces Institute of Pathology, sinonasal mucosal melanomas account for only 1% of all melanomas and for 0.6% to 4% of all tumors of the nasal cavity and paranasal sinuses (9). Conley reported in his series that melanoma involved up to 6.7% of all sinonasal malignancies (10). Ganly et al. have studied 1307 patients with malignant skull base tumors and found that 4% of them had sinonasal melanoma (11). Lewis and Martin have studied the incidence of malignant melanoma of the nasal cavity in Ugandan Africans and found that this disease accounted for 2.6% of all cases of melanoma (12). In Denmark, melanoma of the upper aerodigestive tract accounts for 0.8% of all melanoma cases and for 8% of all head and neck melanomas (13). Japanese have the highest rate of mucosal melanomas compared to Caucasians (4:1 relative rate), most commonly in the oral cavity (14). Interestingly, the incidence of cutaneous melanoma among Japanese is lower than Caucasians. The mean age of presentation of sinonasal melanoma is 65 with a range of 50 to 80 years (15–17). The gender distribution of sinonasal melanomas shows a slight male predominance (18–22). Ganly et al. have reported that of a total of 53 patients with anterior skull base melanoma, 70% were males and 30% females (11). Manolidis and Donald reviewed 172 cases of nasal melanomas and reported that 57% of the cases were males and 43% females (23). Female patients with cutaneous melanoma tend to have a better prognosis than men. However, similar outcomes occurred for males and females with sinonasal melanoma in regard to long-term survival (24–26). The main factor involved in the development of cutaneous melanoma is exposure to sunlight and ultraviolate radiation, whereas the etiology of the ultraviolate lightprotected mucosal melanoma remains unknown. Holmstrom and Lund have suggested that prolonged occupational exposure to formaldehyde may cause significant mucosal irritation eventually causing paranasal malignant melanoma (27). Similarly, Thompson et al. reported formaldehyde exposure in 9 of 115 patients (7.8%) in whom a work history could be elicited (9). The majority of these patients were painters, furniture makers, construction workers, and laundry workers. A joint Danish–Finnish–Swedish case-referent investigation initiated in 1977 studied the connection between nasal and sinonasal cancer and various occupational exposures. The authors found that formaldehyde was evenly distributed among cases with different tumors of paranasal origin (28). The nasal cavity is the most common origin of head and neck mucosal melanomas (55–79%), followed by the oral cavity (29–32). The lateral wall of the nasal cavity and the inferior turbinate are the most common origin of melanoma
Melanoma is a malignancy of ectodermal origin that involves the skin in the vast majority of the cases. The disease is classically divided according to the site of origin of the primary tumor, i.e., cutaneous, noncutaneous, and unknown primary melanoma. Noncutaneous melanomas are infrequent and may be found in the retina, genitourinary tract, anus, and upper aerodigestive tract. The most frequent origin of noncutaneous melanoma is the eye (5.3%) followed by melanomas of an unknown origin (2.2%) and mucosal melanomas (1.3%) (1). Mucosal melanomas of head and neck origin can arise in the oral cavity, oropharynx, nasal cavity, and paranasal sinuses. Melanoma of the sinonasal cavities is a rare neoplasm that can involve various compartments of the respiratory mucosa. The first description of mucosal melanoma was by Weber in 1856 (2). Thirteen years later, Lucke reported the first resection of a nasal mucosal melanoma in a 52-year-old man with “melanotic sarcoma” (3). One of the first reports of sinonasal melanoma is attributed to Viennois, who described the surgical extirpation of “polype melanique du nez” infiltrating the globe (4). In 1885, Lincoln made the first report in the English literature of a “melanosarcoma” arising in the nasal antrum and treated with galvanocauterization (5). Since then more than 1200 cases of mucosal melanoma have been reported in the English literature, one-third of them originating in the sinonasal area. Significant progress has been made during the last decade in the understanding of the biology of melanomas, and its sensitivity to chemotherapy, radiotherapy, and immunotherapy (6). However, due to the low prevalence of sinonasal melanoma, the pathophysiology of the disease as well as the role of radiotherapy and immunotherapy in its treatment still remains controversial. In this chapter, we review the current literature on paranasal and nasal mucosal melanomas and present the results of the latest studies on the epidemiology, pathology, staging, and treatment of patients with this disease.
INCIDENCE AND EPIDEMIOLOGY The incidence of melanoma has been increasing 4% to 6% each year since 1973, a greater rate than any other human cancer in the United States (7). It was estimated that nearly one in 75 persons will develop melanoma during their lifetime (8). The main factor believed to be involved in the significant rise in the incidence of cutaneous melanoma is exposure to sunlight and ultraviolate radiation. Unlike melanoma of the skin, mucosal melanoma did not show any increase in incidence during the last decade, suggesting that a distinct pathophysiologic mechanism is associated with this tumor. An analysis of the National Cancer Data Base performed throughout a period of 19 years (1985–1994) showed 459
460
Gil et al. Table 1
Site of Origin of Sinonasal Melanomas
Site Nasal cavity Nasal NOS Septum Lateral wall Inferior turbinate Middle turbinate Turbinate NOS Floor of nose Subtotal Paranasal sinuses Maxillary sinus Ethmoid sinus Frontal sinus Sinus NOS Subtotal Total a
Manolidis and Donald (23) Number (%) 159 (48) 44 (13) 44 (13) 12 (3.7) 7 (2.1) 12 (3.7) 5 (1.5) 283 (88) 32 (9.8) 8 (2.4) 2 (0.6) 21 (6.4) 63 (18.2) 328 (100)
Nandapalan et al. (34) Number (%)
61 (37) 10 (6.7) 34 (6.1) 43 (26) 2 (1.2) 150 (92) 5 (3) 8 (4.6)
13 (7.9) 163 (100)
a
Thompson et al. (9) Number (%) 34 (44) 20 (26) 10 (13)
64 (84) 3 (4)
9 (12.5) 12 (16) 72 (100)
In other 39 patients the specific compartment was not specified. Abbreviation: NOS, not otherwise specific.
of the nasal mucosa. In the sinuses, the exact site of origin of melanoma is difficult to identify since most tumors are diagnosed at advanced stage and frequently infiltrate multiple compartments. The most common site of origin for melanoma of the sinuses is the maxillary sinus, followed by the ethmoid and sphenoid sinuses (23,26). Melanomas of the frontal sinus, as well as the middle turbinate, superior turbinate, and cribriform plate are very rare (33). Table 1 shows the sites of origin of sinonasal melanomas among 563 patients reviewed in three different studies (9,23,34). Advance stage tumors most frequently involve multiple compartments. For example, in a recent multicenter study, Ganly et al. have reported that 50% of the patients with skull base melanomas had tumors infiltrating the cribriform plate, 34% had periorbital invasion, 26% orbital invasion, and 17% dural invasion (11). The majority of patients with sinonasal melanomas present without regional or distant metastases, at the time of diagnosis. Positive lymph nodes are found in 4% to 18% of the patients at the time of diagnosis (9,19,20,22). The incidence of lymph node involvement in patients with mucosal melanoma of the oral cavity and oropharynx is 4.7 times higher than in patients with sinonasal melanoma, due to the dense lymphatic drainage of the oral cavity and oropharyngeal mucosa (29). Distant metastases of sinonasal melanoma are rare at presentation. For example, Harrison et al. reported that none of the 40 patients in their series had distant metastases at presentation (19). In those patients who develop distant metastases, the most common sites are the lungs, bone, liver, brain, and skin (29).
PATHOLOGY The primary cell of origin of melanoma is the melanocyte, which can be found in the nasal and paranasal mucosa. These melanocytes have migrated as neuroectodermal derivatives and embedded in the endodermally derived respiratory mucosa (35). It is estimated that the distribution of melanocytes in the respiratory mucosa is 1500 cells/mm2 , which is less than two-thirds of that found in the skin (36). Mucosal melanosis is defined as a benign pigmented lesion characterized by pigmentation of basal keratinocytes
with normal or slightly increased number of melanocytes (37). Association between mucosal melanosis and increased incidents of nasal or paranasal melanoma was previously suggested by several authors (12,38,39). A preexisting pigmentation of the sinonasal mucosa is seen in less than 10% of the patients with mucosal malignant melanoma (40). In accordance with this finding, Thompson et al. have found preexisting melanosis in only 8% of the patients with sinonasal melanoma (9). The low rate of melanosis in patients with mucosal melanoma indicates that most sinonasal melanomas arise de novo. The majority of melanomas of the respiratory mucosa are large and polypoid with a median thickness of 9 mm, significantly thicker than melanoma of the oral cavity (41). Similarly, Thompson et al. reported a mean thickness of 7.2 mm (range 2–19 mm) and size of 24 mm (range 5–65 mm) (9). Infrequently, mucosal melanoma in situ can be identified after the biopsy of a suspected lesion (Fig. 1). The authors found no correlation between survival and tumor thickness. Lee et al. have found that depth of invasion > 7 mm is an independent factor for a poor prognosis in patients with mucosal melanomas of the head and neck (42). Melanoma has a notorious tendency to mimic other tumors, and in the sinonasal mucosa it can be easily confused with other tumors, which occur relatively more commonly in this region. The challenge in making the diagnosis of sinonasal melanoma is more significant in case of amelanotic melanoma and in the presence of ulceration. A study by the Armed Forces Institute of Pathology has previously shown that more than two-thirds of the cases of sinonasal melanoma are initially misclassified as other neoplasm on initial pathologic evaluation (9). Histologically, mucosal melanoma cells may have different characteristics including small cells, spindle cells, epithelioid cells, or rarely pleomorphic cells (21,43). Spindle cell melanoma appears sarcomatous, and is composed of cells with eosinophilic cytoplasm and nuclei that may vary in shape and number. Epithelioid-type melanoma is characterized by large cells with eosinophilic cytoplasm and acentric nucleus. Although melanomas display morphologic diversity, undifferentiated small round cells or polygonal cells are the most prominent cells found in sinonasal melanomas
Chapter 32: Melanoma of the Nasal Cavity and Paranasal Sinuses
461
(A)
Figure 1 Mucosal melanoma in situ. The photomicrograph shows mucosal melanoma confined to the sinonasal respiratory epithelium. The neoplastic melanocytes are three to four times larger than the benign surrounding cells. Convoluted nuclear membranes, as well as prominent nucleoli with increased nuclear to cytoplasmic ratio, can also be identified. 100× H&E staining.
(41,44). A pseudopapillary growth pattern is found in up to 25% of patients with sinonasal melanomas, but not in melanoma of the oral cavity (Fig. 2) (21,41). The most significant factor in establishing the diagnosis of melanoma is melanin production and the appearance of junctional changes (Fig. 3). Melanin pigment is found in two-thirds of the cases of sinonasal melanoma and should be considered when confronted by a sinonasal myxoid tumor with melanin (9). Such an example is melanoma botryoides, a polypoid tumor that contains small amounts of melanin with a botryoid or myxoid pattern. Amelanotic tumors show similar biologic behavior and prognosis as melanotic nasal melanomas, but are far more difficult to diagnose compared with conventional melanomas (45). A variant of melanoma, which does not contain melanin, is desmoplastic melanoma (46). Rarely found in sinonasal melanoma, these cells are comprised of amelanotic, poorly circumscribed fascicles and bundles of spindle cells with hyperchromatic nuclei (47). Desmoplastic melanoma is a neurotrophic tumor, which frequently expresses aberrant p53 protein (46). This variant of amelanotic malignant melanoma is difficult to differentiate from other soft tissue tumors of the nasal cavity such as esthesioneuroblastoma, sarcoma, spindle cell carcinoma, and malignant peripheral nerve sheath tumors (48). Melanoma of the sinonasal mucosa often demonstrates deep invasion, necrosis, and vascular invasion. These characteristics are well-established predictors of reduced survival in cutaneous melanomas and melanomas of the head and neck (41). Prasad et al. have shown that 60% of paranasal melanomas are detected in an advanced stage, presenting with significant infiltration of skeletal muscles, cartilage, and bone at the time of surgery (41). Rarely nasal mucosal melanomas may show bone and osteoid formation (9,49). Osteoid and metaplastic bone formation may be caused by repeated trauma, mesenchymal metaplasia, and secondary reaction to bone invasion and introduction of bone formation from the surrounding tissues (9).
(B)
(C)
Figure 2 Nonmelanotic nasopharyngeal mucosal melanoma. (A) Small round blue cells grow radially outward from central vessels demonstrating the pseudopapillary architecture one may frequently see in mucosal melanoma. (B) A photomicrograph of the specimen showing malignant cells arranged in a pavement-like sheet, with no evidence of pigmentation. Large cherry red macronucleoli in the center of the cell nucleus are very characteristic, but not pathognomonic for melanoma. 600× H&E staining. (C) Immunohistochemical staining for S-100, in both the nucleus and cytoplasm are indicative of malignant melanoma.
462
Gil et al.
or paranasal origin. Thompson et al. studied 115 cases of sinonasal melanoma and found positive staining for S-100 in 91% of the cases and HMB-45 in 76% of the cases (9). Tyrosinase antigens were expressed in 77.7% of the cases. Table 2 shows the different antigens expressed in various tumors of the paranasal sinuses including malignant melanoma. Electron microscopy can be utilized as an adjunct for the diagnosis of melanoma in borderline cases and is very specific in identifying premelanosomes, a subcellular organelle present in melanomas. Wright and Heenan identified a subclass of premelanosomes with a high propensity to metastasize (56), however, these premelanosomes are not found in all cases of mucosal melanoma (19,57). Since melanoma of the paranasal sinuses accounts for less than 7% of neoplasms involving this anatomic compartment, a proper different diagnosis workup is crucial for determination of the management of the disease. The microscopic differential diagnosis of sinonasal melanoma can be divided into three groups: small round blue cell tumors, pleomorphic cell tumors, and spindle cell tumors (9). The first group includes tumors such as olfactory neuroblastoma, primitive neuroectodermal tumor, Ewing sarcoma, melanocytic neuroectodermal tumor of infancy, pituitary adenoma, lymphoma, plasmacytoma, small cell neuroendocrine carcinoma, and mesenchymal chondrosarcoma. The pleomorphic cells group includes sinonasal undifferentiated carcinoma, anaplastic large cell lymphoma, and rhabdomyosarcoma. The group of tumors characterized by spindle cells includes malignant peripheral nerve sheath tumor, fibrosarcoma, malignant fibrous histiocytoma, and synovial sarcoma (9,44,58–62). Olfactory esthesioneuroblastoma can show similar morphology to melanoma; however, this tumor frequently shows neurofibrillary background, Homer–Wright rosettes, and focal sustentacular cell pattern even in high-grade tumors (Hyman grade II–IV). Furthermore, most esthesioneuroblastoma shows immunocytochemical reactions to neuronspecific enolase and chromogranin and negative staining to vimentin. As described earlier, mucosal melanoma frequently lack melanin pigment and therefore can be indistinguishable from other high-grade tumors such as sinonasal undifferentiated carcinoma, undifferentiated nasopharyngeal carcinoma, poorly differentiated nonkeratinizing squamous cell carcinoma, small cell carcinoma, and anaplastic large cell lymphoma (63). Fortunately, most epithelial tumors show strong cytokeratin immunostaining and fail to express S-100 protein. Anaplastic large cell lymphoma does not stain for cytokeratin
Figure 3 Melanin pigmented cells in malignant melanoma of the sinonasal mucosa. The most important factor in establishing the diagnosis of melanoma is the appearance of malignant cells containing melanin. Unlike carcinoma, these cells have visible apparent spaces between their cytoplasmic boarders.
Due to the complicated differential diagnosis of this tumor, immunocytochemical staining is frequently required to establish the diagnosis of melanoma of the paranasal sinuses, particularly in cases of amelanotic variants. Melanoma stains positive for vimentin, HMB-45, and S-100 protein. In contrast to cutaneous melanomas, paranasal melanoma is negatively stained to synaptophysin and actin leukocytic common antigen (50,51). Interestingly, staining for P-97, which is frequently positive in melanoma of the esophageal mucosa, is not found in melanomas originating in the paranasal sinuses or nasal mucosa (52,53). Other antigens, which are specifically stained in paranasal melanomas, are KC-2, SK46, and KBA-62 (54). Other melanoma-associated antigens found in the nasal cavity are tyrosinase (T311), D5, A103 (anti-melan-A/MART-1), and TRP-1 (39). In a study of 44 sinonasal melanomas, Prasad et al. have found that all tumors were positive for tyrosinase, 98% for HMB-45, 95% for S-100 protein, and 91% for D5 (55). They concluded that tyrosinase is the most sensitive marker for melanomas of nasal
Table 2
SCC SNUC ONB SCNUC MMM T/NK ML RMS PNET/EWS ∗
Histologic and Antigenic Characteristics of Malignant Melanoma and Other Undifferentiated Tumors Originating in the Nasal and Paranasal Sinuses CK
NSE
CG
SYN
S100
HMB
LCA
CD56
CD99
VIM
DES
Myf4
+ + − + − − − R+
− V + + − − − V
− − V + − − − −
− − V + − − − V
− − +∗ + + − − V
− − − − + − − −
− − − − − V − −
− − − − − + − −
− − − − − − − +
− − − − + V + +
− − − − − − + −
− − − − − − + −
Positive in the peripherally situated sustentacular cells. Abbreviations: SCC, squamous cell carcinoma; SNUC, sinonasal undifferentiated carcinoma; SCUNC, small-cell undifferentiated neuroendocrine carcinoma; MMM, mucosal malignant melanoma; T/NK ML, nasal-type T/natural killer-cell lymphoma; RMS, rhabdomyosarcoma; PNET/EWS, primitive (peripheral) neuroectodermal tumor/extraosseous Ewing sarcoma; CK, cytokeratin; NSE, neuron-specific enolase; CG, chromogranin; SYN, synaptophysin; HMB, HMB 45 (as well as other melanocytic markers [melan A]); CD99, Ewing marker; VIM, vimentin; DES, desmin; (+), positive; (−) negative; v, variably positive; R+, rarely positive. Source: From Ref. 63.
Chapter 32: Melanoma of the Nasal Cavity and Paranasal Sinuses
463
but express CD30 and anaplastic large cell lymphoma kinase proteins. Malignant peripheral nerve sheath tumors may also express S-100 protein, complicating the pathologic diagnosis of melanoma. Malignant peripheral nerve sheath tumor does not express other antigens, such as HMB-45, which are frequently found in sinonasal melanoma.
CLINICAL PRESENTATION AND FINDINGS The clinical symptoms in patients with paranasal sinuses melanoma are frequently nonspecific and include pain, malaise, and weight loss. Other symptoms may be directly associated with the location of the primary tumor. Melanomas originating in the nasal septum may cause irritation and be visible to the patients (64). This may explain the observation that 75% of the patients with nasal mucosal melanoma are diagnosed early with a clinically localized disease, in comparison to those with paranasal sinuses melanoma. Most patients with paranasal melanoma suffer from nasal discharge, recurrent epistaxis, or nasal obstruction (85–90% of cases). Thompson et al. reported that 52 of the 115 patients with melanoma of the sinuses had epistaxis, 42 had a visible mass, and 32 had obstructive symptoms (9). Pain (20%) and a visible facial mass (9%) occur in more advanced disease stage (20). Signs representing orbital involvement are proptosis, ophthalmoplegia, decreased visual fields, and monocular blindness and are associated with poor prognosis (11,42). Most skin or oral cavity melanomas are more likely to be discovered by the patient or by the primary health care physician upon routine examination, whereas sinonasal melanomas are inaccessible to self-examination and are routinely diagnosed at an advanced stage. The duration of the symptoms depends on the biologic behavior of the disease. In case of a slowly growing tumor of paranasal origin, airway obstruction may develop slowly and the disease, which is obscured from the patient and the physician, may develop months or years before the diagnosis is established (19). Most authors report mean symptoms duration of 2 to 8 months (13,20,31,42,65). Holdcraft and Gallagher have reported that 50% of their patients had suffered from symptoms for 1 to 4 months before diagnosis, and 30% had symptoms for 6 to 24 months (18). Thompson et al. reported a mean duration of 8.2 months ranging between 2 weeks and 8 years (9). Evaluation of patients with suspected sinonasal melanoma should include complete history and physical evaluation with emphasis on the head and neck. Fiberoptic evaluation of the paranasal sinuses and upper aerodigestive tracts is indicated in all patients, in order to evaluate the tumor extent and potential for resection (Fig. 4). Radiologic evaluation should always include both computed tomography (CT) and magnetic resonance imaging (MRI) of the head, neck, and paranasal sinuses for evaluation of bony and soft tissue involvement, respectively (66). Patients should be evaluated for involvement of cranial nerve, orbit, skull base, dural or brain infiltration using both CT and MRI. Although perineural spread of disease occurs more commonly with squamous cell carcinoma and adenoid cystic carcinoma of the paranasal sinuses, malignant melanoma must also be included in this differential diagnosis, particularly if the patient’s pathology is known to be desmoplastic melanoma. Imaging in patient with malignant melanoma of the paranasal sinuses should focus on the likely potential for perineural spread. In a recent study of nine patients with melanomas of the facial skin and paranasal sinuses (including five desmoplastic melanomas) with symptomatic cranial
Figure 4 Endoscopic photograph of a mucosal melanoma of the paranasal sinuses. Fiberoptic examination of the paranasal sinuses and upper aerodigestive tract is indicated in all patients prior to surgery, in order to evaluate the tumor extent and the presence of a second primary. The photograph shows mucosal melanoma arising from the sinonasal mucosa (arrow).
neuropathy, MR imaging demonstrated post gadolinium enhancement of the trigeminal nerve in all nine cases and of other cranial nerves in five cases (67). Other findings included abnormal contrast enhancement and soft tissue thickening in the cavernous sinus, Meckel cave, and/or the cisternal segment of the trigeminal nerve. Suspicious neck metastases can be evaluated with CT, MRI, or ultrasound. Ultrasound-guided fine-needle biopsy can be utilized if indicated. Sentinel lymph node biopsy, which is frequently used for lymphatic mapping of cutaneous melanomas, is not a common practice for sinonasal disease (68). Chest radiograph should be performed for evaluation of lung metastases. Chest CT, liver ultrasound, and bone scan should be performed if metastatic disease is suspected. Patients with mucosal melanomas can be evaluated for the presence of metastases using positron emission tomography (PET). Goerres et al. have found that all mucosal melanomas of the head and neck can be visualized using FDG PET (66). Large lesions with a nodular growth are better demonstrated than lesions with a superficial mucosal spread. Similarly, lesions originating in the nasal vestibule are more challenging to identify than those in the posterior sinonasal complex, due to nonspecific uptake in the skin and muscles of the mouth. Due to the high yield in staging metastatic disease, utilizing metabolic PET imaging can replace staging techniques employing multiple imaging modalities (i.e., chest x-ray, neck and liver ultrasound, total body CT, and bone mapping) (69). Metastatic cutaneous melanoma to the paranasal sinuses is very rare and account for 1% of the patients with cutaneous melanoma (44,70). Nevertheless, despite its rarity, full workup should be performed to exclude isolated metastasis of melanoma of the skin to the paranasal sinuses.
STAGING There is no formal staging system for mucosal or sinonasal melanoma. The TNM classification of melanomas is only
464
Gil et al.
available for skin and ocular lesions. The staging of cutaneous melanoma is based on the staging system revised by the American Joint Committee on Cancer (AJCC) (71). In this system, the prognosis of patients with localized disease (T1: tumors < 1.0 mm or T2: 1–2 mm in thickness) is good, whereas for patients with melanomas >2.0 mm in thickness, a worse survival rate is expected (T3, T4). Patients with localized disease and no regional or distant metastases are classified as having stage I or II disease. In patients with regional metastases (stage III), tumor burden is expressed as the number of positive nodes (N1 for a single node, N2 for 2–3 nodes and N3 if ≥4 lymph nodes are involved). In patients with distant metastatic disease (stage IV), the sites of metastases determine outcome and M classification is graded from a–c accordingly (i.e., skin, lung, and visceral, respectively). As with cutaneous melanoma, the outcome of mucosal melanoma initially depends on the stage at presentation (42). Unfortunately, because of an absence of histologic landmarks identifiable as a papillary and reticular dermis in the respiratory mucosa, the AJCC cutaneous classification system cannot be applied for this disease. Furthermore, sinonasal melanomas are frequently polypoid rather than being deeply invasive and tumor thickness cannot accurately predict the prognosis (9). An alternative system for classification of sinonasal melanoma is the AJCC staging criteria for nasal and paranasal epithelial tumors (72). As described in detail in chapter 29 “Squamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses,” this staging system considers the extent of the primary lesion (stage I–IV) and the presence of regional metastases (stage III–IV) or distant metastases (stage IV) as main prognostic indicators for survival. In the absence of a conventional classification for sinonasal melanomas, few authors have adopted the AJCC staging system for nasal and paranasal epithelial tumors for sinonasal mucosal melanoma (20,29,73). One must take into consideration that the AJCC TNM classification applies only to histologically confirmed carcinomas, and it was not validated for melanoma. The most common system used for staging of mucosal melanoma of the head and neck was first suggested by Ballantyne (74). In this classification, a stage I disease represents tumors confined to the primary site, stage II denotes the existence of positive regional lymph nodes, and stage III indicates a distant metastatic disease. This system does not take into consideration the size and extent of the disease process, but has been used repeatedly for classification of sinonasal melanomas and melanomas of the upper aerodigestive tract (26,30,75). Since regional and metastatic disease is not common during initial diagnosis, most patients are characterized in the stage I group. Clearly, the main drawback of this staging system is its lack of ability to differentiate between patients with localized disease, and those with advanced tumors and poor prognosis (i.e., orbital infiltration or intracranial extension). Thompson et al. have studied a group of 115 patients with sinonasal melanomas in an attempt to develop a validated staging system by incorporating features of size, site, regional and distant metastases into a single staging system (9). The T classification of this staging system separates between tumors localized to a single anatomic compartment (T1) and those involving more than one anatomic level (T2). The N classification accounts for the absence (N0) or presence (N1) of lymph node metastases (whether ipsilateral, bilateral or contralateral). Patients with T1 and T2 disease in the absence of regional or distant metastases are grouped into disease stage I and II, respectively. A stage III disease accounts for patients with any T, N1, and M0, whereas pa-
Table 3 Proposed Staging for Sinonasal Tract and Nasopharynx Mucosal Malignant Melanoma Nasal cavity, paranasal sinuses, and nasopharynx histopathology staging Primary tumor T1 T2 Regional lymph node N1 Distant metastasis M1 Stage grouping Stage I Stage II Stage III Stage IV
Single anatomic site Two or more anatomic sites Any lymph node metastasis Distant metastasis T1, N0 M0 T2, N0 M0 Any T, any N, M1 Any T, any N, M1
T, primary tumor. TX, primary tumor cannot be assessed. T0, no evidence of primary tumor. T1, tumor limited to a single anatomic site. A single anatomic site is defined as one of the following: nasal cavity, maxillary sinus, frontal sinus, ethmoid sinus, sphenoid sinus, nasopharynx. Subsites, such as septum, lateral wall, turbinate, nasal floor, or nasal vestibule are not separately considered. T2, tumor involving more than one anatomic site. More than one anatomic site is defined by tumor involvement of more than one anatomic site (although not subsite) as cited above, including any extension into subcutaneous tissues, skin, palate, pterygoid plate, floor, wall, or apex of the orbit, cribriform plate, infratemporal fossa, dura, brain, middle cranial fossa, cranial nerves, clivus. N, regional lymph nodes (cervical lymph nodes). NX, regional lymph nodes cannot be assessed. N0, no regional lymph node metastasis. N1, metastasis in regional lymph node(s) of any size, whether ipsilateral, bilateral, or contralateral (midline nodes are considered ipsilateral nodes). M, distant metastasis. MX, distant metastasis cannot be assessed. M0, no distant metastasis. M1, distant metastasis. pTNM, pathological classification. The pT, pN, and pM categories correspond to the T, N, and M categories. From a practical standpoint, documentation of metastatic disease (lymph node or distant) is based on findings within 90 days peri-diagnosis (ie, a lymph node is the initial presentation and a mucosal primary is documented within 3 months; a STMMM is diagnosed and then CT, MR or other studies are performed over the ensuing 6 weeks and identify metastatic disease). pN0, histologic examination of a selective neck dissection specimen will ordinarily include 6 or more lymph nodes. Histologic examination of a radical or modified radical neck dissection specimen will ordinarily include 10 or more lymph nodes. If the lymph nodes are negative, but the number ordinarily examined is not met, classify as pN0. Source: Adapted from Ref. 9.
tients with distant metastases are classified as having a stage IV disease. The TNM classification suggested by Thompson et al. predicted patients’ outcome based on the anatomic site of involvement and metastatic disease (Table 3). Prasad et al. at Memorial Sloan–Kettering Cancer Center suggested further microscopic classification of stage I (lymph node-negative) sinonasal melanoma (76). Their microstaging system was performed according to disease invasion into three compartments: level I, melanoma in situ; level II, invasion into the lamina propria; and level III, invasion into deep tissue (i.e., skeletal muscles, bone, or cartilage). Kaplan– Meier analysis showed significant difference in 5-year disease
Chapter 32: Melanoma of the Nasal Cavity and Paranasal Sinuses
specific survival of patients with level I (75%), level II (52%), and level III (23%).
TREATMENT There is a general consensus that surgery remains the treatment of choice for mucosal melanoma of sinonasal origin (75). A recent report of the National Cancer Institute and the Center for Disease Control has demonstrated an absolute gain in overall survival of melanoma patients during the last decade (77). The improvement in survival of cutaneous melanoma patients, despite the increase in incidence, is attributed to early detection and improvements in therapy. The mode of therapy for sinonasal melanoma awaits further evaluation due to lack of prospective studies and objective data for the benefit of adjuvant modalities such as postoperative radiotherapy, immunotherapy, and chemotherapy.
Surgery In the absence of distant metastases, complete tumor extirpation is the mainstay of treatment for malignant melanoma of the paranasal sinuses and nasal mucosa. Although negative margins after surgical resection is reported in over 85% of patients, it is frequently impossible to achieve en bloc tumor resection, and 75% of the patients will eventually develop local recurrence (11). Two possible explanations for the high recurrence rate are (i) presence of a multifocal disease and (ii) submucosal lymphatic spread of melanoma cells in the respiratory mucosa, which is not clinically or radiographically apparent (26). Freedman et al. suggested multicentricity as a main factor predicting local recurrence after surgery (20). Thus wide surgical resection, without unnecessary compromise of function and cosmesis is essential.
(A)
(B)
465
The route of spread of tumors originating in the anterior skull base and paranasal sinuses is determined by the complex anatomy of the craniofacial compartments. A tumor arising in the ethmoid sinus or paranasal cavity may invade laterally to the orbit, inferiorly to the maxillary antrum and palate, posteriorly to the nasopharynx and pterygopalatine fossa (PPF), and superiorly to the dura, brain, or cavernous sinus. The recent improvement of endoscopic technology now allows for the resection of benign neoplasms or early malignant neoplasms (78). However, for sinonasal melanoma, an open approach is more suitable to allow extirpation of tumors in an en bloc fashion and with wide margins, in the opinion of the authors. The type of surgery is planned according to the extent of the tumor. For small tumors involving the nasal septum, resection of the tumor along with the perichondrium and septal cartilage may be performed via lateral rhinotomy incision (Fig. 5). However, frequently these tumors also infiltrate adjacent structures such as the hard palate, ethmoid sinuses, and medial maxillary walls. In these cases a unilateral or bilateral medial maxillectomy is performed via a lateral rhinotomy incision. Conventional exposure of the infra- and suprastructure of the maxilla is achieved via a lateral rhinotomy with lip split or subciliary extension as indicated (Fig. 6). Tumors infiltrating the cribriform plate and fovea ethmoidalis are safely accessed via the craniofacial or subcranial approach (11). These approaches offer wide exposure of the tumor, allowing resection of the intracranial and extracranial extensions of the tumor. Massive orbital involvement or orbital apex infiltration requires orbital exenteration. In this case, orbital exenteration is performed with radical maxillectomy or craniofacial resection, as determined by the tumors extension. Combinations of the craniofacial approach with transorbital and middle fossa
(C)
Figure 5 Nonmelanotic melanoma of the nasal septum. (A) Endoscopic photograph of the lesion shows a nonmelanotic nasal septal lesion (asterisk, nasal septum; arrow, lateral nasal wall). (B) A preoperative coronal CT scan showing a left nasal septal lesion. (C) An intraoperative picture demonstrating exposure of the lesion. The type of surgery is planned according to the extent of the tumor. For small tumors involving the nasal septum, resection of the tumor along with the perichondrium and septal cartilage can be performed via lateral rhinotomy incision.
466
(A)
Gil et al.
(B)
(C)
Figure 6 Mucosal melanoma of the left maxillary sinus. (A) A preoperative coronal CT scan showing a hyperdense mass in the right maxillary antrum. (B) An intraoperative picture demonstrating exposure of the lesion. En bloc resection of the tumor along with a radical maxillectomy was achieved via the transfacial approach (lateral rhinotomy with lip split). (C) The excised specimen showing the melanotic lesion extending into the maxillary antrum.
approaches were described in detail by Shah et al. for malignant tumors of the anterior skull base and paranasal sinuses (79). A combined facial translocation approach was also described and safely used by Hao et al. for malignant tumors of the paranasal cavity (80,81). Following resection of the primary tumor, the surgical margins should be evaluated for residual disease using multiple permanent sections at the periphery of resection, sampling bone, mucosa, soft tissue and other tissue as indicated. Although complete tumor resection should be the goal of care for sinonasal melanoma, a recent analysis of 53 patients with skull base melanoma has shown no survival benefit of negative margins in this anatomic area (11). Dural and anterior skull base reconstruction is required after craniofacial resection. Dural reconstruction is performed principally with pericranial, galeal, temporalis fascia, or fascia lata grafts. Bovine pericardium can also be utilized for reconstruction. Fibrin glue is used in order to provide additional protection against cerebrospinal fluid leak. Reconstruction of the medio-orbital walls is not typically performed. If indicated, a split calvarial bone graft, a fascia lata sling, or 3dimensional titanium mesh covered by pericranium are used for reconstruction of the orbit. A temporalis muscle flap and a split-thickness skin graft to cover the orbital socket can be used after orbital exenteration. In cases of a radical maxillectomy with or without orbital exenteration, a lateral thigh free flap or a rectus abdominis musculocutaneous free flap may be utilized to obliterate this large defect and to support the obturator. Since neck lymph nodes are rarely encountered in cases of sinonasal melanoma, neck dissection should only be performed if regional metastases are identified, based on clinical or radiologic evaluation. The postoperative complication rate following surgical resection of malignant skull base tumors is 36% (11).
Among patients treated with craniofacial resection for excision of sinonasal melanomas, the postoperative mortality is 6% and major postoperative complications occur in 26% of the patients.
Radiation The use of radiation therapy for treatment of melanoma is controversial. Despite the long-standing debate regarding the radiosensitivity of melanoma reported in the past, there has been a significant increase in the use of adjuvant radiation therapy for the treatment of this disease (1). Both clinical and basic science data indicate that melanoma cells have the ability to repair cellular damage, providing resistance to conventional fractionated radiation therapy (82). It was therefore speculated that hypofractionation or high-dose-perfractionation (HDPF) therapy will give more effective radiation treatment to these patients. Several nonrandomized, retrospective studies have reported improved locoregional control rates of patients with high-risk cutaneous melanoma of the head and neck using conventional or hypofractionation adjuvant radiation compared with surgery alone (83,84). Moreover, Raben et al. have reported 70% local control rate in 10 patients treated with the high-dose-per-fractionation regimen following surgical resection of head and neck malignant melanomas, with minimal morbidity (85). However, no change in overall survival was found in this study following hypofractionation adjuvant radiotherapy due to the development of disseminated disease. In patients with mucosal melanoma, Patel et al. reported no advantage of postoperative radiotherapy compared to surgery alone (26). In contrast, Ganly et al. reported that postoperative radiation therapy was an independent positive predictor of overall, disease-specific, and recurrence-free survival on a multivariate analysis of patients with skull base melanoma (11). Patients treated with surgery and postoperative radiotherapy
Chapter 32: Melanoma of the Nasal Cavity and Paranasal Sinuses Table 4
Melanoma of the Nasal and Paranasal Sinuses and Survival
Study
Year
N
5-year survival
Impact of radiotherapy on survival
Holdcraft & Gallagher (18) Freedman et al. (20) Eneroth & Lundberg (91) Harrison (19) Trapp et al. (16) ∗ Gilligan et al. (22) Kingdom and Kaplan (15) Harbo et al. (92) Brandwein et al. (58) Lund VJ et al. (77) Owens et al. (85) Thompson et al. (9) Patel et al. (26) Nakaya et al. (93). Bridger et al. (72) Ganly et al. (11)
1968 1973 1975 1976 1987 1991 1995 1997 1997 1999 2003 2003 2002 2004 2005 2006
39 56 24 40 17 28 17 25 36 72 11 115 35 16 27 53
10% 30% 17% 27.5% 25 18% 20% 24% 36% 28% 33% 42.6% 47% 31.8% 43% 24%
– NS – – – – Prolonged – – NS NS NS NS NS – >2-fold increase
∗
467
Primary radiotherapy. Abbreviation: NS, nonsignificant.
had 39% 3-year overall survival, compared with 18% in patients treated with surgery alone. Freedman et al. have reported no survival benefit for patients with paranasal and nasal cavity melanoma treated with surgery and adjuvant radiotherapy compared with those treated with surgery alone (20). In the same study, it was reported that none of the 18 patients treated with radiation alone survived after 5 years. Owens et al. recently reported a retrospective evaluation of patients with mucosal melanoma (23% of them with sinonasal disease) treated with surgery alone, surgery and adjuvant radiotherapy, or surgery and biochemotherapy, with or without adjuvant radiotherapy (86). Radiation therapy was generally used as an adjuvant therapy for patients with extensive disease. The patients with sinonasal tumors received 6000 cGy in 30 fractions, while those with oral lesions received 3000 cGy in 5 fractions. Biochemotherapy (cisplatin, vinblastine, and dacarbazine, with or without the addition of interferon alfa-2b and interleukin 2) was used almost exclusively in patients who had recurrent disease or distant metastases. The addition of radiotherapy tended to decrease the rate of local failure, but did not prevent distant metastases or improve overall survival. Biochemotherapy regimens used for metastatic or recurrent disease had no significant impact on survival. An evaluation of the impact of postoperative radiotherapy on local control and survival of patients with head and neck mucosal melanoma was reported by Temam et al. at the Institut Gustave–Roussy (Villejuif) (29). Sixty-nine patients with local disease were managed by surgery without postoperative radiotherapy, two-thirds of whom had sinonasal disease. The study suggested that postoperative radiotherapy increased local control in patients with small tumors, but did not impact survival. Most reports of definitive radiation therapy for mucosal melanomas involve small series of patients. Gilligan and Slevin reported one of the largest series of melanomas of the paranasal sinuses and nasal cavity treated with radiation alone (22). Complete response was achieved in 79% of the 28 patients included in the study, with relatively low treatment morbidity. In this study, the overall 3- and 5-year survival was 49% and 18%, respectively (Table 4). Stern and Guillamondegui at MD Anderson Cancer Center reported two of the
five patients alive and disease free, 5 years after radiotherapy alone (31). At the Princess Margaret Hospital in Toronto, Canada, Harwood and Cummings reported 50% local control rate at 6 months to 4.2 years after primary radiotherapy (n = 10 patients) (87). Trotti and Peters reviewed a series of reports using radiotherapy alone for mucosal head and neck melanoma and concluded that 50% to 75% of the patients had documented complete response, with long-term control in 50% to 66% (88). They concluded that in view of the poor results of radical surgery, radiation should be seriously considered as the initial treatment of choice for primary mucosal melanomas of the head and neck. Albertsson et al. reported the result of hyperfractionation radiation in combination with cisplatinum (89). Three of four patients treated for local recurrence achieved local control. Radiation may be also appropriate as a primary treatment for patients with unresectable disease, elderly patients, patients with high surgical risk or palliative therapy (90). External beam radiation may achieve local control, relieve pain, and decrease tumor mass compression on vital structures including cranial nerves, orbit, airway or brain. Radiation therapy has the potential for complications, especially if applied to the cranial base. Severe morbidity associated with radiation of the anterior skull base includes osteoradionecrosis, encephalomalacia, optic neuropathy and retinopathy (94). Radiation therapy has also been associated with a decreasing quality of life in patients with skull base malignancies (95). Using heavy particle radiation sources (i.e., proton or carbon ions), as well as accurate delivery using intensity-modulated radiation therapy, may be beneficial in enhancing therapeutic outcomes and reducing complication rates. Fast neutron therapy was used to treat primary, recurrent, or metastatic cutaneous and mucosal melanoma in 48 patients, showing complete regression in 71%, with a 9% incidence of local recurrence (96). The median survival was 14.5 months and complications, including fibrosis and necrosis, occurred in 22% of patients. In another series, Linstadt et al. reported local control in two of the six patients treated with neon ions for melanoma located in a variety of sites including paranasal sinus (97). Promising results were found in a dose escalation study using carbon ions, reporting 100% 5-year
Gil et al.
local control rate in five patients with mucosal melanoma (98).
Chemotherapy and Immunotherapy Chemotherapy is regarded today mainly as palliative treatment for recurrent, metastatic, or inoperable disease. A variety of chemotherapeutic agents, both as adjuvant or neoadjuvant therapy, have been used for treatment of distant disease based on treatment of patients with cutaneous melanomas. Cisplatinum, dacarbazine, and vinca alkaloids are some of the chemotherapeutic agents used for systemic treatment. Unfortunately, most authors reported no advantage for chemotherapy in patients with distant metastatic melanoma originating from a head and neck mucosal primary (13,25,30,99). An emergent therapeutic strategy for treatment of systemic disease is the use of immunotherapy or biologic response modifiers with or without chemotherapy. The most frequently used biologic response modifiers are bacillus Calmette–Guerin (BCG) vaccine, interleukin-2, and interferon alpha-2b (IFN alpha-2b). In 2002, the National Comprehensive Cancer Network committee had recommended for high-risk patients with localized cutaneous melanomas greater than 4.0 mm in thickness to participate in adjuvant therapy clinical trials, including treatment with high-dose adjuvant IFN alpha-2b versus observation (100). This recommendation was based on two studies performed by the Eastern Cooperative Oncology Group (trial 1684 and 1690); these were randomized controlled studies of IFN alpha-2b administered at maximum tolerated doses versus observation (101). Another study showed a relapse-free survival advantage for high-dose interferon with no difference in overall survival (102). Legha et al. evaluated the use of concurrent biochemotherapy including cisplatin, vinblastine, and dacarbazine in combination with IFN-alpha and interleukin-2 in patients with metastatic melanoma (103). Among the 53 patients treated in this study, 21% achieved a complete response and 43% achieved a partial response. The median survival was 11.8 months. The toxicity reported was severe myelosuppression, nausea, vomiting, and hypotension that required inpatient care and support. There were no treatmentrelated deaths. An anecdotal report that evaluated the utility of hormonal therapy with tamoxifen for palliative treatment of patients with sinonasal melanoma reported good response in all three patients treated with the drug (91). Since there are no trials for adjuvant immunotherapy treatment of sinonasal mucosal melanoma, the decisions regarding the use of adjuvant therapy for these patients should be made on an individual basis, extrapolating from available data from the cutaneous adjuvant trials, and after discussion with the patient, including an explanation of the adjuvant treatment clinical results and anticipated morbidity.
OUTCOME AND PROGNOSIS Of all patients with mucosal melanomas of the head and neck, those with disease involving the paranasal sinuses have the poorest outcome. Notably, tumors isolated to the nasal cavity are associated with the best prognosis of head and neck mucosal melanoma. Patients with sinonasal melanomas have an interposed course of disease with multiple local or regional recurrences, followed by distant metastases and ultimately death from disseminated disease. Head and neck mucosal melanomas have the lowest 5-year survival rate compared to cutaneous and ocular melanomas (32%, 75%, and 80%, respectively) (1). The estimated 10-year survival rate of these
patients is 7% (104). In spite of the development of new surgical approaches and novel therapeutic agents, there was no significant increase in the overall survival of patients with sinonasal and nasal cavity melanomas. A large cohort study performed by the Armed Forces Institute of Pathology reported no influence of treatment modality on overall survival (9). Among the 115 patients participated in this study, there was no statistical significant difference between patients managed by surgery alone, surgery with chemotherapy, surgery with radiotherapy or surgery with combined therapy. After a mean follow-up of almost 14 years, 35% of the patients were alive or have died without evidence of disease. Owens et al. reported 14.3 months’ average interval to failure in patients with sinonasal tumors (range, 3–36 months), compared to 31.6 months in patients with oral or oropharyngeal lesions (range, 3–147 months) (86). For patients with sinonasal disease, the overall survival rates at 3 and 5 years were 50% and 33%, respectively. A recent report of a large cohort of patients with sinonasal melanoma undergoing craniofacial resection accumulated from multiple international institutions reported a 3-year disease-free and overall survival of 28% and 29%, respectively (Fig. 7) (11). Orbital involvement was found to be an independent predictor of overall and disease specific survival. For comparison, the overall 5-year survival of patients with all malignant skull base tumors treated by the same group of surgeons was 54%, including esthesioneuroblastoma (64%) and squamous cell carcinoma (50%) (105). Ganly et al. demonstrated a threefold increase in overall and disease-free survival following adjuvant conventional radiation therapy (11). The risk of recurrence was fourfold in patients not receiving radiotherapy. As a result of incomplete data regarding precise staging, dose or delivery of radiation, and previous treatment, such retrospective studies must be interpreted with caution. Harrison et al. reported 3-, 5-, and 10-year disease-specific survival of 47%, 28% and 8%, respectively with surgery alone (19). Similar results were reported by Freedman and colleagues who found 46% and 31% survival at 3 and 5 years among their group of patients treated with surgery and postoperative radiotherapy (20).
1.0 3-Year OS (28.2%) 3-Year DSS (29.7%) 3-Year RFS (25.5%)
0.9 0.8 Proportion surviving
468
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
0
10
20
30 40 Follow-up time, mo
50
60
Figure 7 Survival analysis of patients with malignant melanoma of the paranasal sinuses. Three-year overall survival (OS), disease-specific survival (DSS), and recurrence-free survival (RFS) for craniofacial resection for malignant melanoma invading the skull base. Source: From Ref. 11.
Chapter 32: Melanoma of the Nasal Cavity and Paranasal Sinuses
LOCAL
NECK
17%
6%
3%
8% 12%
3%
14%
DISTANT Figure 8 Patterns of recurrence in sinonasal melanomas. No recurrence was found in 11 of the 35 patients (31.3%), isolated locoregional recurrence in 11 of 35 (31.3%), isolated distant failure in 4 of 35 (11.4%), and local and/or regional recurrence with distant failure in 9 of 35 patients (25.7%). Source: From Ref. 26.
Local failure is a significant cause of mortality in patients with sinonasal melanoma. Disease recurrence is common within the first 2 years after treatment and late recurrence may be seen even after 5 years of follow-up. Freedman et al. reported a series of 56 patients with sinonasal melanoma, 34 of them (60%) had recurrence in the primary site and 29% underwent salvage resection of the tumor (20). Patel et al. reported local failure rate of 50%, nodal failure rate of 20%, and distant failure rate of 40% (Fig. 8) (26). Only 6% of the patients in their study were eligible for salvage therapy. Patients with cutaneous melanoma and negative lymph nodes at presentation have a 5-year survival rate of 80% compared with 30% for those with positive nodes. In contrast, patients with mucosal melanoma of the head and neck have 27% and 19% 5-year survival for N0 and N+ disease, respectively (26,92). Similar results were found by Yii et al. who reported 26% 5-year survival rate for patients with localized disease, compared to 0% 5-year survival for patients with regional or distant metastases (32). Temam et al. reported that pathologic neck stage did not influence the overall survival of patients with mucosal melanomas of the head and neck (29). In most patients, local recurrence is an ominous sign for ongoing distant disease. Although rare at presentation, 37% of the patients will ultimately fail at distant sites, more than two-thirds of whom will also develop local or regional disease. Stern and Guillamondegui reported that 89% of patients with local recurrence also develop disseminated disease (31). The most common site of distant metastases is the lung and brain (33% and 14%, respectively). The median survival period from the time of detection of distant metastases to death is 7.1 months (26).
469
Although there are reports of anecdotal cases in which the disease remained dormant for long period of time, most patients with sinonasal melanoma die of distant metastases with or without local recurrence. The unpredictability of the clinical behavior of this tumor is commonly characterized by cases with a prolonged clinical course despite repeated local recurrences and regional lymph node metastases (104). Thus, there is a lifelong risk of recurrence for all patients with sinonasal mucosal melanoma. For this reason, lifelong surveillance of these patients is required. The value of salvage surgery for local recurrences and regional lymph node metastases cannot be predicted based on clinical and pathologic parameters alone. It is plausible that the course of disease is based on interactions between the primary tumor and the host’s immune system, but the exact mechanism of such interaction is currently not evident.
CONCLUSIONS Melanoma of the nasal cavity and paranasal sinuses is an uncommon malignancy arising in the respiratory sinonasal mucosa. The majority of cases originate in the nasal cavity followed by the paranasal sinuses. Whereas most head and neck melanomas are more likely to be discovered by the patient or during routine physical examination, sinonasal melanomas are not accessible to self-examination and are often diagnosed late, resulting in poor survival. Melanoma has the tendency to mimic other tumors pathologically and can lead to initial misdiagnosis. Due to the complicated differential diagnosis of this tumor, immunocytochemical staining is frequently required to establish the diagnosis of melanoma, particularly in cases of amelanotic variants. Surgery is considered the treatment of choice for primary mucosal melanoma of the sinonasal cavities. Due to the nature of this disease, it is challenging to achieve complete tumor resection and the majority of those patients will eventually develop local recurrence. The high recurrence rate of sinonasal melanoma is secondary to a high incidence of multifocal disease and the presence of submucosal lymphatic spread of tumor cells. Thus, all lesions require wide surgical resection, attempting to minimize unnecessary compromise of function and cosmesis. Of the patients with mucosal melanomas of the head and neck, those with disease involving the paranasal sinuses have the poorest outcome, whereas tumors isolated to the nasal cavity are associated with a better prognosis. Local failure is a significant cause of mortality in these patients and is clearly associated with high rate of nodal recurrence and distant metastases. Most patients with local recurrence are not amended to curative salvage therapy. The use of radiation therapy for treatment of melanoma is controversial. Recent studies indicate that the addition of adjuvant radiotherapy tends to decrease the rate of local failure, but has no significant impact on overall survival. Primary radiotherapy or chemotherapy is currently employed as palliative treatment of recurrent, inoperable or metastatic disease, or for patients with unacceptable surgical risk. Unfortunately, most reports showed no survival advantage for chemotherapy in patients with disseminated disease. The utility of other treatment modalities such as immunotherapy or biochemotherapy, as well as heavy particle radiation sources and intensity-modulated radiation therapy , awaits further evaluation. Further studies are required to determine the advantage of postoperative radiotherapy for treatment of patients with sinonasal melanoma. Nevertheless, it is advisable to add hypofractionation radiotherapy after surgery,
470
Gil et al.
especially for treatment of sinonasal melanomas, which historically has been associated with a high incidence of local recurrence after surgery alone.
ACKNOWLEDGMENT We thank Diana L. Carlson, MD, Department of Pathology, Memorial Sloan–Kettering Cancer Center, for providing the original pathological photographs for this chapter.
REFERENCES 1. Chang AE, Karnell LH, Menck HR. The National Cancer Data Base report on cutaneous and noncutaneous melanoma: A summary of 84,836 cases from the past decade. The American College of Surgeons Commission on Cancer and the American Cancer Society. Cancer. 1998;83(8):1664–1678. 2. Weber CO. Chirurgische Ehrfahrungen und Untersuchungen, nebst zahlreichen Beobachtungen aus der chirurgischen Klinik und dem evangelischen Krahkenhause zu Bonn. Berlin: G. Reimer, 1859. 3. Lucke A. Die Lehre von den Geschwulsten in anatomischer und klinischer Beziehung in Handbuch d. allg. u. spec. chir. Erlangen. 1869:244. 4. Viennois L. Osteotomie du nez (Obs. II - Polype melanique du nez (melanosarcome) ayant detruit la paroi interne de l’orbite et refoulant le globe de l’oeil en avant: abaissement du nez comme operation preliminaire; ablation du polyp, pyohemie; mort huit jours apres). Lyon Med. 1872;11:8–12. 5. Lincoln RP. A case of melano-sarcoma of the nose cured by galvanocauterization. N Y Med J. 1885;42:406–407. 6. Chin L. The genetics of malignant melanoma: Lessons from mouse and man. Nat Rev Cancer. 2003;3(8):559–570. 7. Wingo PA, Ries LA, Rosenberg HM, et al. Cancer incidence and mortality, 1973–1995: A report card for the U.S. Cancer. 1998;82(6):1197–1207. 8. Rigel DS. Malignant melanoma: Incidence issues and their effect on diagnosis and treatment in the 1990s. Mayo Clin Proc. 1997;72(4):367–371. 9. Thompson LD, Wieneke JA, Miettinen M, et al. Sinonasal tract and nasopharyngeal melanomas: A clinicopathologic study of 115 cases with a proposed staging system. Am J Surg Pathol. 2003;27(5):594–611. 10. Conley JJ. Melanoma of the Head and Neck. 1st ed. New York, NY: Georg Thiem Verlag, 1990:154–178. 11. Ganly I, Patel SG, Singh B, et al. Craniofacial resection for malignant melanoma of the skull base: Report of an international collaborative study. Arch Otolaryngol Head Neck Surg. 2006;132(1):73–78. 12. Lewis MG, Martin JA. Malignant melanoma of the nasal cavity in Ugandan Africans. Relationship of ectopic pigmentation. Cancer. 1967;20(10):1699–1705. 13. Andersen LJ, Berthelsen A, Hansen HS. Malignant melanoma of the upper respiratory tract and the oral cavity. J Otolaryngol. 1992;21(3):180–185. 14. Takagi M, Ishikawa G, Mori W. Primary malignant melanoma of the oral cavity in Japan. With special reference to mucosal melanosis. Cancer. 1974;34(2):358–370. 15. Kingdom TT, Kaplan MJ. Mucosal melanoma of the nasal cavity and paranasal sinuses. Head Neck. 1995;17(3):184–189. 16. Trapp TK, Fu YS, Calcaterra TC. Melanoma of the nasal and paranasal sinus mucosa. Arch Otolaryngol Head Neck Surg. 1987;113(10):1086–1089. 17. Blatchford SJ, Koopmann CF Jr, Coulthard SW. Mucosal melanoma of the head and neck. Laryngoscope. 1986;96(9 Pt 1):929–934. 18. Holdcraft J, Gallagher JC. Malignant melanomas of the nasal and paranasal sinus mucosa. Ann Otol Rhinol Laryngol. 1969;78(1):5–20.
19. Harrison DF. Malignant melanomata arising in the nasal mucous membrane. J Laryngol Otol. 1976;90(11):993–1005. 20. Freedman HM, DeSanto LW, Devine KD, et al. Malignant melanoma of the nasal cavity and paranasal sinuses. Arch Otolaryngol. 1973;97(4):322–325. 21. Matias C, Corde J, Soares J. Primary malignant melanoma of the nasal cavity: A clinicopathologic study of nine cases. J Surg Oncol. 1988;39(1):29–32. 22. Gilligan D, Slevin NJ. Radical radiotherapy for 28 cases of mucosal melanoma in the nasal cavity and sinuses. Br J Radiol. 1991;64(768):1147–1150. 23. Manolidis S, Donald PJ. Malignant mucosal melanoma of the head and neck: Review of the literature and report of 14 patients. Cancer. 1997;80(8):1373–1386. 24. Moore ES, Martin H. Melanoma of the upper respiratory tract and oral cavity. Cancer. 1955;8(6):1167–1176. 25. Panje WR, Moran WJ. Melanoma of the upper aerodigestive tract: A review of 21 cases. Head Neck Surg. 1986;8(4):309– 312. 26. Patel SG, Prasad ML, Escrig M, et al. Primary mucosal malignant melanoma of the head and neck. Head Neck. 2002;24(3):247– 257. 27. Holmstrom M, Lund VJ. Malignant melanomas of the nasal cavity after occupational exposure to formaldehyde. Br J Ind Med. 1991;48(1):9–11. 28. Hernberg S, Westerholm P, Schultz-Larsen K, et al. Nasal and sinonasal cancer. Connection with occupational exposures in Denmark, Finland and Sweden. Scand J Work Environ Health. 1983;9(4):315–326. 29. Temam S, Mamelle G, Marandas P, et al. Postoperative radiotherapy for primary mucosal melanoma of the head and neck. Cancer. 2005;103(2):313–319. 30. Guzzo M, Grandi C, Licitra L, et al. Mucosal malignant melanoma of head and neck: Forty-eight cases treated at Istituto Nazionale Tumori of Milan. Eur J Surg Oncol. 1993;19(4):316– 319. 31. Stern SJ, Guillamondegui OM. Mucosal melanoma of the head and neck. Head Neck. 1991;13(1):22–27. 32. Yii NW, Eisen T, Nicolson M, et al. Mucosal malignant melanoma of the head and neck: The Marsden experience over half a century. Clin Oncol (R Coll Radiol). 2003;15(4):199–204. 33. Ravid JM, Esteves JA. Malignant melanoma of the nose and paranasal sinuses and juvenile melanoma of the nose. Arch Otolaryngol. 1960;72:431–444. 34. Nandapalan V, Roland NJ, Helliwell TR, et al. Mucosal melanoma of the head and neck. Clin Otolaryngol Allied Sci. 1998;23(2):107–116. 35. Zak FG, Lawson W. The presence of melanocytes in the nasal cavity. Ann Otol Rhinol Laryngol. 1974;83(4):515–519. 36. Szabo G. The number of melanocytes in human epidermis. Br Med J. 1954;4869:1016–1017. 37. Mannone F, De Giorgi V, Cattaneo A, et al. Dermoscopic features of mucosal melanosis. Dermatol Surg. 2004;30(8):1118–1123. 38. Cove H. Melanosis, melanocytic hyperplasia, and primary malignant melanoma of the nasal cavity. Cancer. 1979;44(4):1424– 1433. 39. Hofbauer GF, Boni R, Simmen D, et al. Histological, immunological and molecular features of a nasal mucosa primary melanoma associated with nasal melanosis. Melanoma Res. 2002;12(1):77–82. 40. Bradley PJ. Primary malignant mucosal melanoma of the head and neck. Curr Opin Otolaryngol Head Neck Surg. 2006;14(2):100–104. 41. Prasad ML, Busam KJ, Patel SG, et al. Clinicopathologic differences in malignant melanoma arising in oral squamous and sinonasal respiratory mucosa of the upper aerodigestive tract. Arch Pathol Lab Med. 2003;127(8):997–1002. 42. Lee SP, Shimizu KT, Tran LM, et al. Mucosal melanoma of the head and neck: The impact of local control on survival. Laryngoscope. 1994;104(2):121–126. 43. Crawford RI, Tron VA, Ma R, et al. Sinonasal malignant melanoma—A clinicopathologic analysis of 18 cases. Melanoma Res. 1995;5(4):261–265.
Chapter 32: Melanoma of the Nasal Cavity and Paranasal Sinuses 44. Franquemont DW, Mills SE. Sinonasal malignant melanoma. A clinicopathologic and immunohistochemical study of 14 cases. Am J Clin Pathol. 1991;96(6):689–697. 45. Chetty R, Slavin JL, Pitson GA, et al. Melanoma botryoides: A distinctive myxoid pattern of sino-nasal malignant melanoma. Histopathology. 1994;24(4):377–379. 46. Prasad ML, Patel SG, Busam KJ. Primary mucosal desmoplastic melanoma of the head and neck. Head Neck. 2004;26(4):373– 377. 47. Batsakis JG, Regezi JA, Solomon AR, et al. The pathology of head and neck tumors: Mucosal melanomas, part 13. Head Neck Surg. 1982;4(5):404–418. 48. Kilpatrick SE, White WL, Browne JD. Desmoplastic malignant melanoma of the oral mucosa. An underrecognized diagnostic pitfall. Cancer. 1996;78(3):383–389. 49. Hoorweg JJ, Loftus BM, Hilgers FJ. Osteoid and bone formation in a nasal mucosal melanoma and its metastasis. Histopathology. 1997;31(5):465–468. 50. Ordonez NG, Ji XL, Hickey RC. Comparison of HMB-45 monoclonal antibody and S-100 protein in the immunohistochemical diagnosis of melanoma. Am J Clin Pathol. 1988;90(4):385–390. 51. Gown AM, Vogel AM, Hoak D, et al. Monoclonal antibodies specific for melanocytic tumors distinguish subpopulations of melanocytes. Am J Pathol. 1986;123(2):195–203. 52. Chello M, Marchese AR, Panza A, et al. Primary malignant melanoma of the oesophagus with a left atrial metastasis. Thorax. 1993;48(2):185–186. 53. Mohandas KM, Santhi Swaroop V, Dhir V, et al. Primary malignant melanoma of the esophagus palliated with endoscopic laser therapy. Endoscopy. 1993;25(8):551. 54. Uno Y, Saito R, Kanatani M, et al. Malignant melanoma of the nasal cavity. Auris Nasus Larynx. 1994;21(4):248–252. 55. Prasad ML, Jungbluth AA, Iversen K, et al. Expression of melanocytic differentiation markers in malignant melanomas of the oral and sinonasal mucosa. Am J Surg Pathol. 2001;25(6):782–787. 56. Wright JL, Heenan PJ. Electron microscopy findings in malignant melanoma of nose. ORL J Otorhinolaryngol Relat Spec. 1975;37(4):233–236. 57. Henzen-Logmans SC, Meijer CJ, Ruiter DJ, et al. Diagnostic application of panels of antibodies in mucosal melanomas of the head and neck. Cancer. 1988;61(4):702–711. 58. Brandwein MS, Rothstein A, Lawson W, et al. Sinonasal melanoma. A clinicopathologic study of 25 cases and literature meta-analysis. Arch Otolaryngol Head Neck Surg. 1997;123(3):290–296. 59. Lentsch EJ, Myers JN. Melanoma of the head and neck: Current concepts in diagnosis and management. Laryngoscope. 2001;111(7):1209–1222. 60. Mills SE, Fechner RE. “Undifferentiated” neoplasms of the sinonasal region: Differential diagnosis based on clinical, light microscopic, immunohistochemical, and ultrastructural features. Semin Diagn Pathol. 1989;6(4):316–328. 61. Nakhleh RE, Wick MR, Rocamora A, et al. Morphologic diversity in malignant melanomas. Am J Clin Pathol. 1990;93(6):731– 740. 62. Wick MR, Stanley SJ, Swanson PE. Immunohistochemical diagnosis of sinonasal melanoma, carcinoma, and neuroblastoma with monoclonal antibodies HMB-45 and anti-synaptophysin. Arch Pathol Lab Med. 1988;112(6):616–620. 63. Ejaz A, Wenig BM. Sinonasal undifferentiated carcinoma: Clinical and pathologic features and a discussion on classification, cellular differentiation, and differential diagnosis. Adv Anat Pathol. 2005;12(3):134–143. 64. Kuwabara H, Takeda J. Malignant melanoma of the lacrimal sac with surrounding melanosis. Arch Pathol Lab Med. 1997;121(5):517–519. 65. Hoyt DJ, Jordan T, Fisher SR. Mucosal melanoma of the head and neck. Arch Otolaryngol Head Neck Surg. 1989;115(9):1096– 1099. 66. Hammersmith SM, Terk MR, Jeffrey PB, et al. Magnetic resonance imaging of nasopharyngeal and paranasal sinus melanoma. Magn Reson Imaging. 1990;8(3):245–253.
471
67. Chang PC, Fischbein NJ, McCalmont TH, et al. Perineural spread of malignant melanoma of the head and neck: Clinical and imaging features. Am J Neuroradiol. 2004;25(1):5–11. 68. Civantos FJ, Moffat FL, Goodwin WJ. Lymphatic mapping and sentinel lymphadenectomy for 106 head and neck lesions: Contrasts between oral cavity and cutaneous malignancy. Laryngoscope. 2006;112(3)(Pt 2)(Suppl 109):1–15. 69. Boni R, Boni RA, Steinert H, et al. Staging of metastatic melanoma by whole-body positron emission tomography using 2-fluorine-18-fluoro-2-deoxy-D-glucose. Br J Dermatol. 1995;132(4):556–562. 70. Billings KR, Wang MB, Sercarz JA, et al. Clinical and pathologic distinction between primary and metastatic mucosal melanoma of the head and neck. Otolaryngol Head Neck Surg. 1995;112(6):700–706. 71. Balch CM, Buzaid AC, Soong SJ, et al. Final version of the American Joint Committee on Cancer staging system for cutaneous melanoma. J Clin Oncol. 2001;19(16):3635–3648. 72. American Joint Committee on Cancer. Head and Neck Sites. 6th ed. New York, NY: Springer, 2002:17–87. 73. Bridger AG, Smee D, Baldwin MA, et al. Experience with mucosal melanoma of the nose and paranasal sinuses. ANZ J Surg. 2005;75(4):192–197. 74. Ballantyne AJ. Malignant melanoma of the skin of the head and neck. An analysis of 405 cases. Am J Surg. 1970;120(4):425– 431. 75. Mendenhall WM, Amdur RJ, Hinerman RW, et al. Head and neck mucosal melanoma. Am J Clin Oncol. 2005;28(6):626–630. 76. Prasad ML, Patel SG, Huvos AG, et al. Primary mucosal melanoma of the head and neck: A proposal for microstaging localized, Stage I (lymph node-negative) tumors. Cancer. 2004;100(8):1657–1664. 77. Jemal A, Clegg LX, Ward E, et al. Annual report to the nation on the status of cancer, 1975–2001, with a special feature regarding survival. Cancer. 2004;101(1):3–27. 78. Lund VJ, Howard DJ, Harding L, et al. Management options and survival in malignant melanoma of the sinonasal mucosa. Laryngoscope. 1999;109(2 Pt 1):208–211. 79. Shah J. Nasal cavity and paranasal sinuses. In: Mosbey, ed. Head and Neck Surgery and Oncology. New York, NY: Elsevier, 2003:57–93. 80. Hao SP, Pan WL, Chang CN, et al. The use of the facial translocation technique in the management of tumors of the paranasal sinuses and skull base. Otolaryngol Head Neck Surg. 2003;128(4):571–575. 81. Bilsky MH, Bentz B, Vitaz T, et al. Craniofacial resection for cranial base malignancies involving the infratemporal fossa. Neurosurgery. 2005;57:339–347. 82. Bentzen SM, Overgaard J, Thames HD, et al. Clinical radiobiology of malignant melanoma. Radiother Oncol. 1989;16(3):169– 182. 83. Ang KK, Byers RM, Peters LJ, et al. Regional radiotherapy as adjuvant treatment for head and neck malignant melanoma. Preliminary results. Arch Otolaryngol Head Neck Surg. 1990;116(2):169–172. 84. Stevens G, Thompson JF, Firth I, et al. Locally advanced melanoma: Results of postoperative hypofractionated radiation therapy. Cancer. 2000;88(1):88–94. 85. Raben A, Zelefsky M, Harrison LB. High dose per fractionation, short course irradiation for mucosal melanoma of the head and neck. Presented at the 4th International Head and Neck Society Meeting. Toronto, Canada, June 15, 1996. 86. Owens JM, Roberts DB, Myers JN. The role of postoperative adjuvant radiation therapy in the treatment of mucosal melanomas of the head and neck region. Arch Otolaryngol Head Neck Surg. 2003;129(8):864–868. 87. Harwood AR, Cummings BJ. Radiotherapy for mucosal melanomas. Int J Radiat Oncol Biol Phys. 1982;8(7):1121–1126. 88. Trotti A, Peters LJ. Role of radiotherapy in the primary management of mucosal melanoma of the head and neck. Semin Surg Oncol. 1993;9(3):246–250. 89. Albertsson M, Tennvall J, Andersson T, et al. Malignant melanoma of the nasal cavity and nasopharynx treated with
472
90.
91.
92. 93.
94.
95.
96. 97.
98.
Gil et al. cisplatin and accelerated hyperfractionated radiation. Melanoma Res. 1992;2(2):101–104. Wada H, Nemoto K, Ogawa Y, et al. A multi-institutional retrospective analysis of external radiotherapy for mucosal melanoma of the head and neck in Northern Japan. Int J Radiat Oncol Biol Phys. 2004;59(2):495–500. Seo W, Ogasawara H, Sakagami M. Chemohormonal therapy for malignant melanomas of the nasal and paranasal mucosa. Rhinology. 1997;35(1):19-21. Shah JP, Huvos AG, Strong EW. Mucosal melanomas of the head and neck. Am J Surg. 1977;134(4):531–535. Harbo G, Grau C, Bundgaard T, et al. Cancer of the nasal cavity and paranasal sinuses. A clinico-pathological study of 277 patients. Acta Oncol. 1997;36(1):45–50. Garrott H, O’Day J. Optic neuropathy secondary to radiotherapy for nasal melanoma. Clin Experiment Ophthalmol. 2004;32(3):330–333. Gil Z, Abergel A, Spektor S, et al. Quality of life following surgery for anterior skull base tumors. Arch Otolaryngol Head Neck Surg. 2003;129(12):1303–1309. Blake PR, Catterall M, Errington RD. Treatment of malignant melanoma by fast neutrons. Br J Surg. 1985;72(7):517–519. Linstadt DE, Castro JR, Phillips TL. Neon ion radiotherapy: Results of the phase I/II clinical trial. Int J Radiat Oncol Biol Phys. 1991;20(4):761–769. Mizoe JE, Tsujii H, Kamada T, et al. Organizing Committee for the Working Group for Head-And-Neck Cancer. Dose escalation
99.
100. 101.
102.
103.
104.
105.
study of carbon ion radiotherapy for locally advanced headand-neck cancer. Int J Radiat Oncol Biol Phys. 2004;60(2):358– 364. Jacquillat C, Khayat D, Banzet P, et al. Chemotherapy by fotemustine in cerebral metastases of disseminated malignant melanoma. Cancer Chemother Pharmacol. 1990;25(4):263–266. National Comprehensive Cancer Network, Inc Version 1.2005, 2005. www.nccn.org Kirkwood JM, Strawderman MH, Ernstoff MS, et al. Interferon alfa-2b adjuvant therapy of high-risk resected cutaneous melanoma: The Eastern Cooperative Oncology Group Trial EST 1684. J Clin Oncol. 1996;14(1):7–17. Kirkwood JM, Ibrahim JG, Sondak VK, et al. High- and low-dose interferon alfa-2b in high-risk melanoma: First analysis of intergroup trial E1690/S9111/C9190. J Clin Oncol. 2000;18(12):2444– 2458. Legha SS, Ring S, Eton O, et al. Development of a biochemotherapy regimen with concurrent administration of cisplatin, vinblastine, dacarbazine, interferon alfa, and interleukin2 for patients with metastatic melanoma. J Clin Oncol. 1998;16(5):1752–1759. Eneroth CM, Lundberg C. Mucosal malignant melanomas of the head and neck with special reference to cases having a prolonged clinical course. Acta Otolaryngol. 1975;80(5–6):452–458. Patel SG, Singh B, Polluri A, et al. Craniofacial surgery for malignant skull base tumors: Report of an international collaborative study. Cancer. 2003;98(6):1179–1187.
33 Sarcomas of the Skull Base Katherine A. Thornton and Robert S. Benjamin
INCIDENCE AND EPIDEMIOLOGY
ment and are treated at specialty centers (4). The most comprehensive data addressing this issue in soft tissue sarcoma come from Sweden, where Gustafson and colleagues analyzed the quality of treatment in a population-based series of 375 patients with primary soft tissue sarcomas arising in the extremities (n = 329) or the trunk (n = 46) (5). Comparisons were made between patients referred to a specialty soft tissue tumor center before surgery (n = 195), those referred after surgery (n = 102), and those not referred for treatment of the primary tumor (n = 78). The total number of operations for the primary tumor was 1.4 times higher in the patients not referred and 1.7 times higher in the patients referred after surgery than in patients referred before surgery. Of greatest significance, however, was the finding that the local recurrence rate was 2.4 times higher in the patients not referred and 1.3 times higher in the patients referred after surgery than in patients referred to a specialty soft tissue tumor center before any manipulation of their tumor. These findings support the principle of centralizing treatment of these rare tumors, which frequently require complex multimodality therapy. While these data were collected on patients with sarcomas of the extremities and trunk, they should apply with even more certainty to patients with skull base sarcomas, where the expertise is even more limited.
Sarcomas comprise a group of relatively rare, anatomically and histologically diverse neoplasms. These tumors share a common embryologic origin, arising primarily from mesenchymal tissue. The notable exceptions are malignant peripheral nerve sheath tumors and primitive neuroectodermal tumors, also known as Ewing sarcomas, which are believed to arise from neuroectodermal tissue. Despite the fact that the somatic soft tissues account for as much as 75% of total body weight, neoplasms of the soft tissues are comparatively rare, accounting for less than 1% of adult malignancies and 15% of pediatric malignancies. The annual incidence of soft tissue sarcomas in the United States is about 15,000 if those arising in organs are included, according to SEER estimates. For bone sarcomas, the incidence is about 2500 cases. For sarcomas arising in somatic soft tissue, there are only about 8000 new cases, but only about 4% of these or about 300 cases arise in the head and neck. Nonetheless, the overall mortality rate of sarcomas is almost 50% at all sites, and tumors of the skull base are particularly hard to cure due to the inability to perform adequate oncologic surgery due to anatomic constraints. Head and neck sarcomas are uncommon, accounting for less than 1% of head and neck malignancies in adults. In recent large series of 176 (1), 188 (2), and 254 (3) patients with head and neck sarcomas, the most common anatomic sites were the neck (23–38%) and paranasal sinuses (14–30%). Thus, skull base sarcomas represent a small subset of a rare group of tumors. While almost any histologic type of sarcoma can occur in the head and neck, certain tumors are found more commonly at the skull base: hemangiopericytomas of the dura and chordomas. These will be discussed in detail following brief description of various soft tissue and bone sarcomas, any of which can involve the soft tissues or bones of the skull base.
SURGERY Surgical management of sarcomas is no different from that of other tumors at a similar location except that lymph node resection is rarely required. Wide excision of the tumor with a margin of normal tissue is preferred, but anatomic constraints often preclude more than a gross total resection. Details of surgical management are discussed elsewhere in this book. Wide surgical excision with microscopically negative margins is the therapeutic mainstay for all sarcomas, but is rarely achievable at the base of skull. Local recurrence remains a significant problem, with overall rates of local recurrence for all head and neck sarcomas ranging from 14% to 48% (1,2,6), making it especially important to consider appropriately applied principles of sarcoma management with combined modality approaches where appropriate. As with sarcomas elsewhere in the body, biologic behavior is a function of histologic grade, with local recurrence rates ranging from 22% for low-grade head and neck sarcomas to 48% for high-grade lesions (1). Systemic recurrence develops in 12% to 31% of patients despite complete resection (1,6). Overall 5-year survival rates are 45% to 68% (1–3,7).
DIAGNOSIS Methods of diagnosis, imaging, and biopsy for head and neck sarcomas do not differ substantially from those for other skull base tumors. CT and MRI are complementary studies, with CT providing sharper images of cortical bone and MRI showing details of intracranial extension or bone marrow involvement. While fine-needle aspiration may occasionally be definitive, most centers require at least a core biopsy to establish the diagnosis.
TREATMENT Treatment of Sarcoma Patients at Specialty Centers
RADIATION TREATMENT
Recent data on other tumor types have demonstrated improved outcomes for patients who required complex treat-
Based on experience gained in treating extremity sarcomas, adjuvant radiotherapy should be utilized routinely since 473
474
Thornton and Benjamin
there is always doubt as to the adequacy of surgical margins, and, in many cases, the location of the tumor precludes complete excision. Evidence for the benefit of adjuvant radiotherapy for head and neck sarcomas is less plentiful than for extremity lesions; however, Tran and colleagues from the University of California, Los Angeles have shown that local control was 52% with surgery alone versus 90% in head and neck patients treated with combined radiotherapy and surgery (8). Additional evidence from the Princess Margaret Hospital reveals that head and neck STS patients with clear surgical margins or microscopic residuum had similar local failure rates (26% and 30% failure, respectively), provided radiotherapy was administered (9). Indeed, these outcomes for radiotherapy after R0/R1 resections approach those achieved in extremity sarcoma. Data on skull base tumors are less clear. Gil et al. noted that postoperative radiotherapy in their cohort of patients with anterior skull base sarcomas did not affect disease-specific survival, but only 34% of their patient population was irradiated (10). Similarly, Prabhu et al. could not identify benefit in either progression-free survival or overall survival in their patients with skull base sarcomas, 62% of whom received postoperative radiation therapy (11). One strategy to improve outcome in head and neck sarcomas is through the use of preoperative radiotherapy. This approach may have particular advantages in this site because of the smaller volumes of radiotherapy and the lower doses that can be used compared to the postoperative treatment in difficult surgical access locations, especially in the base of skull. Obvious advantages provided relate to the ability to spare critical anatomy such as the optic structures (globes, optic nerves, and the optic chiasm), as well as the brain stem and spinal cord. If for no other reason, the preoperative approach promotes collaboration between the surgical and radiation oncologist, facilitates a complete management plan to be fashioned before any surgical intervention, and maximizes the opportunity to achieve local control even when disease may be resected with a small but planned positive margin against critical unexpendable anatomy, as discussed earlier (12). That said, many head and neck surgeons, including those at MD Anderson, prefer to use radiation only postoperatively. A prospective series of 40 patients with head and neck sarcomas (excluding rhabdomyosarcoma) with adverse selection criteria was managed with preoperative radiotherapy between 1989 and 1999 at the Princess Margaret Hospital. If the four patients with angiosarcoma are excluded (almost always involve the skin, and thus are not relevant to skull base tumors) there were three local relapses in the 36 patients (overall control rate of 92%) (13). This population of head and neck patients included five patients with intracranial extension, one with spinal cord compression, more than half were greater than 5 cm in size (a formidable problem for lesions in this anatomical location), and 85% were deep to the investing fascia. The metastatic relapse-free rate also exceeded 80% in this series, potentially related in part to the smaller overall dimension of sarcomas in this location compared to sarcomas elsewhere. Also the improved local control compared to a previous series of patients treated at the same institution may have contributed to this amelioration because the local control rate in the earlier series was substantially lower and death from concurrent local and metastatic disease was evident (9). Wound complications, assessed by the Canadian trial criteria (14), were also seen with less frequency in this prospective study of head neck lesions (overall rate of 8 of 40 or 20%) (13) than were noted earlier with preoperative
radiotherapy in extremity lesions. This may relate to the greater use of flaps for head and neck reconstruction. At present, useful guidelines for using preoperative radiotherapy in the head and neck are: (i) the need to maximally restrict radiotherapy volumes in some anatomic sites (e.g., close to critical anatomy); (ii) the desire to minimize radiation dose in some situations (e.g., where critical neurological tissues are in close proximity, as in the optic structures); and (iii) a desire not to irradiate new tissues, especially vascular reconstructions vulnerable to the effects of high-dose postoperative radiotherapy. Newer techniques such as IMRT or proton therapy should permit better targeting where appropriate.
CHEMOTHERAPY The activity of chemotherapy for sarcomas was established by treating patients with metastatic disease at various primary sites and with diverse histologic diagnoses. Indeed, considering the diversity of diseases lumped as sarcomas, it is amazing that any drugs have been found to have activity. Lessons from those studies can be applied to patients with skull base sarcomas, but with a number of histology-directed exceptions.
Anthracyclines The single-agent activity of doxorubicin against metastatic soft tissue sarcoma is well established as being in the range of 20% to 40% (15–18). A randomized study in the Southwest Oncology Group demonstrated a steep dose–response curve for doxorubicin in sarcomas, in contrast to other tumors (19). The response rate increased significantly from 17% at 45 mg/m2 to 37% at 75 mg/m2 . Doxorubicin administration is limited acutely by mucositis, especially when given over several days. It is limited chronically by cardiomyopathy that increases after cumulative doses exceeding 400 mg/m2 when given by rapid infusion. With cardioprotective strategies, continuous infusion over 48 to 96 hours or premedication with dexrazoxane, cumulative doses can be doubled, but at the cost of increased mucositis (infusion, particularly >48 hours) or myelosuppression (dexrazoxane). Epirubicin, developed as an active but minimally cardiotoxic analogue of doxorubicin, produced an objective response rate not significantly inferior than that of doxorubicin (18% versus 25%, P = 0.33) in an EORTC RCT of 167 patients receiving equimolar doses (75 mg/m2 ) of the drugs (20). There is some evidence of a dose–response relationship with epirubicin; a dose-escalation study showed response rates of 17%, 44%, and 100% for 140 mg/m2 , 160 mg/m2 , and 180 mg/m2 of epirubicin, respectively (21). Only three patients were entered at the maximum tolerated dose of 180 mg/m2 , and 160 mg/m2 was recommended for routine clinical use; however, the EORTC, in a three-arm randomized study of 334 patients, was unable to demonstrate any benefit from either of two schedules of epirubicin (150 mg/m2 ) compared with doxorubicin (75 mg/m2 ); all regimens produced response rates of 14% to 15% (22). Furthermore, there was considerably more myelosuppression in the two epirubicin arms, with two toxic deaths. Nevertheless, in some areas of Europe, epirubicin is frequently substituted for doxorubicin in high-dose regimens. A number of studies of liposomal anthracyclines have suggested that these agents have lower rates of cardiotoxicity but variable activity (23–27). An EORTC phase II RCT 320 demonstrated low activity for both doxorubicin and liposomal doxorubicin (Doxil), 9% versus 10%, but different
Chapter 33: Sarcomas of the Skull Base
spectrums of toxicity: less myelosuppression but palmer– plantar erythrodysesthesia (grade 3, 20%) as the dose-limiting toxicity with Doxil.
Ifosfamide After the reports of several studies documenting activity ranging from 24% to 67% (28–30), Bramwell and colleagues performed a randomized trial comparing ifosfamide (5 g/m2 by 24-hour infusion) with cyclophosphamide (1.5 g/m2 ) (31). Respective response rates were 18% and 8%, and although the difference was not statistically significant (P = 0.13), responses were seen only in patients failing cyclophosphamide and crossing over to ifosfamide and not the other direction. Further indirect data from several other studies in which ifosfamide and/or cyclophosphamide were added to doxorubicin have provided additional evidence that ifosfamide is a more active analogue than cyclophosphamide, and all sarcoma oncologists agree that ifosfamide is more active (15). Questions about the optimal scheduling of ifosfamide (multiple daily bolus doses versus continuous infusion) have never been satisfactorily resolved and are confounded by dose differences in many studies. Two consecutive phase II studies by investigators at MD Anderson (30) at 8 g/m2 and in Boston (28) at 10 g/m2 evaluated ifosfamide given as a continuous infusion or a 2-hour infusion daily for 4 to 5 consecutive days. Response rates in both groups were higher when intermittent short infusions were used.
High-Dose Ifosfamide Early studies of ifosfamide suggested that there was a dose– response relationship (30), and several groups have documented responses to high-dose ifosfamide in patients not responding to lower doses of the drug (32–35). Nevertheless dose-escalation studies of ifosfamide have produced conflicting results. Doses of 12 g/m2 without and 14 to 18 g/m2 with growth factor support seem achievable and have produced response rates of 33% to 45%, but nephrotoxicity and neurotoxicity are considerable (35–37). Frustaci and colleagues found high-dose ifosfamide to be well tolerated when infused at 1 g/m2 /d over 21 days (38). In 36 patients, they were able to administer up to three cycles of median duration 15 days, producing a response rate of 24%. Myelosuppression was dose limiting, but there was no significant nephrotoxicity or neurotoxicity. Pharmacokinetic data, reported by Cerny and colleagues demonstrated that ifosfamide doses greater than 14 to 16 g/m2 given over 5 days resulted in a relative decrease of the active metabolite phosphoramide mustard, suggesting dose-dependent saturation or inhibition of ifosfamide metabolism (39). Two consecutive phase II studies by investigators at MD Anderson evaluated ifosfamide (14 g/m2 ), given as a 72-hour continuous infusion or a 2-hour infusion for three consecutive days. Respective response rates were 19% and 42% (36).
Combination Chemotherapy In the 1970s and early 1980s before the widespread availability of ifosfamide, most combination chemotherapy regimens were based on doxorubicin and dacarbazine. These are still appropriate for patients over age 65 and those with impaired renal function. The addition of cyclophosphamide and vincristine, which are active against childhood sarcomas, created a regimen called CyVADIC, for which the Southwest Oncology Group reported response rates as high as 59% in patients with metastatic disease (40). Later investigators were unable to reproduce such high response rates with the same regimen, however, and summary data on variants of the CyVADIC
475
regimen revealed an overall response rate of 35% in 2092 patients (41). Most regimens now used for first-line chemotherapy are based on the combination of doxorubicin and ifosfamide. A recent systematic search of the literature (42) found three phase III studies and 16 phase II trials (excluding phase I studies and those recruiting less than 25 patients) in adult STS that used combination regimens including an anthracycline and ifosfamide. Although the response rate varied widely from 25% to 56% in the phase II studies, they were at the lower end of this range in the three phase III studies (43–45). In the Eastern Cooperative Oncology Group (ECOG) study (43) of 178 patients, the response rate was significantly higher for doxorubicin/ifosfamide at 60 mg/m2 and 6 g/m2 than for doxorubicin alone (34% versus 20%, P = 0.03) although median survivals were similar. In an EORTC study of 471 patients, however, there were no significant differences in response rate (28% versus 23%) or median survival for doxorubicin/ifosfamide at 50 mg/m2 and 5 g/m2 versus doxorubicin alone at 75 mg/m2 (44). In contrast, the highest response rates were seen in phase II studies from MD Anderson that exploited the dose–response relationship for the two agents (46). At doxorubicin doses of 75 to 90 mg/m2 and an ifosfamide dose of 10 g/m2 , the response rate for patients with metastatic disease was 62% (46).
Second-Line Chemotherapy Objective response rates have also been very low (3% to 5%) in three phase II studies of gemcitabine (47–49), although the MD Anderson group described a response rate of 18% in 39 patients, if GISTs were excluded (50). It is the only agent commercially available in the United States with significant single-agent activity in previously treated patients. Temozolomide (51–53), raltitrexed (54), irinotecan (55), sargramostim (56), topotecan (57,58), vinorelbine (59), and the taxanes (60–64) seem to have minimal activity in STS, despite their proven value in other tumor types. Of drugs currently in development, trabectidin (Yondelis, ET743), a DNA guanine-specific minor groove-binding agent, seems to have the most potential across the spectrum of sarcomas. Hints of activity in bone and STSs were observed in phase I trials (65,66) and appeared to be confirmed in phase II trials of this agent. Demetri and colleagues reported a response rate of 18% in 34 chemonaive sarcoma patients and 9% in 34 who had received prior chemotherapy. George and colleagues reported a lower progression rate (5%) but a substantial proportion (19%) of patients with minor responses or stable disease (67). Two European trials described response rates of 11% to 12% in previously treated sarcoma patients (68,69). The drug is particularly active against myxoid liposarcoma, where about 80% of patients show benefit (70). It is commercially available in Europe. Occasional severe toxicities, sometimes lethal, seemed to be related to elevated baseline liver function tests. Despite poor levels of activity as single agents, some drugs have been incorporated into nonanthracycline-based salvage regimens. Based on encouraging data in pediatric sarcomas, etoposide has been combined with ifosfamide, although with variable results (71–74). All but one such study produced response rates in the range of 38% to 46%; because ifosfamide given alone has produced up to 67% in phase II studies, however, the contribution of etoposide to these results is difficult to discern. Several investigators, the authors included, do not consider the contribution of etoposide significant in typical adult sarcomas. Combinations of docetaxel with gemcitabine are being evaluated, with preliminary
476
Thornton and Benjamin
reports of encouraging synergistic activity (75–78). In a randomized study conducted by SARC (the Sarcoma Alliance for Research through Collaboration) of equitoxic doses of gemcitabine by timed infusion as a single agent versus the combination of timed infusion gemcitabine plus docetaxel, the combination produced a superior response rate, time to progression, and survival (78).
CHEMOTHERAPY CONSIDERATIONS FOR SPECIFIC HISTOLOGIC SUBTYPES Synovial Sarcoma Rosen and colleagues (79) were the first to suggest that synovial sarcomas were particularly responsive to ifosfamide. They documented three complete responses and nine partial responses in 13 patients (nine of whom had received prior doxorubicin-based chemotherapy) with metastatic synovial sarcomas. These investigators also reported on 14 patients with localized synovial sarcomas who received adjuvant doxorubicin/ifosfamide/cisplatin chemotherapy (80). There was one patient with local recurrence, but the remaining 13 patients (93%) remained disease free at a median follow-up period of 37 months (range, 6–85 months). In a large EORTC phase II trial of 124 patients with advanced STS receiving ifosfamide (12 g/m2 ), the overall response rate for all histological subtypes was 16% (81), but 8 (44%) of 18 patients with synovial sarcoma responded. Edmonson and colleagues described a higher response rate in synovial sarcomas for doxorubicin/ifosfamide than for doxorubicin alone (88% vs. 20%, P = 0.02) in the setting of the ECOG randomized phase III study of multiple histologic subtypes of STS (43). A subsequent ECOG phase II study of doxorubicin/ifosfamide in synovial sarcomas showed five partial responses (42%) in 12 patients; however, the median survival for the whole group was only 11 months, and the trial was closed because of poor accrual (82). In many studies evaluating ifosfamide, including some of the randomized studies, the question of response by histologic subtype has not been specifically addressed. In several studies (28,30,36,81), however, leiomyosarcoma stands out as particularly unresponsive, even when patients with GIST (30,36) are excluded. As many of these patients are young and fit, inclusion of ifosfamide in first-line chemotherapy for metastatic disease seems reasonable. If the circumstances merit adjuvant chemotherapy, an anthracycline/ifosfamide combination is the logical choice for fit patients younger than 65 with good renal function.
Liposarcoma Activation of the PPAR-gamma nuclear receptor stimulates terminal differentiation in preadipocytes. Thioglitazones, used in the treatment of diabetes mellitus, are activating ligands for PPAR-gamma. Biopsies, pre- and post-troglitazone therapy, were obtained in 34 of 49 patients with different types of liposarcomas entering a phase II trial (83). Five of seven (71%) evaluable patients with myxoid/round cell disease exhibited histologic evidence of lineage-appropriate differentiation of liposarcoma cells, whereas only one of three (33%) patients with high-grade pleomorphic disease showed such changes (84). Although this study provides proof-of-concept data, the clinical significance is uncertain as responses to troglitazone were not documented. Trabectidin, as mentioned above, is highly active against myxoid liposarcoma (70). It is significantly less active against other subtypes of liposarcomas and other sarcomas.
Pediatric Sarcomas (Rhabdomyosarcoma and Ewing Sarcoma) in Adults Embryonal rhabdomyosarcomas and the Ewing family of tumors, all seem to be chemosensitive when they occur in the adult age group. Adult patients with these tumors should receive aggressive combination chemotherapy. We prefer to use the basic doxorubicin/ifosfamide regimen, but to add vincristine. Others add vincristine and etoposide, and many prefer alternating regimens similar to those offered to children with the same disease (85,86). Nevertheless, the outcome is likely to be poorer for adults with “pediatric sarcomas” than for pediatric sarcoma patients. Actinomycin-D, inactive in most adult sarcomas, can be effective in both of these tumors, especially rhabdomyosarcomas. Topoisomerase I inhibitors, topotecan and irinotecan, also have definite activity, and are often combined with alkylating agents. These tumors are also more sensitive to radiation than most other sarcomas, and radiation is often an effective substitute for surgery if its use will decrease morbidity substantially.
Osteosarcoma Chemotherapy is usually employed in the neoadjuvant situation, and its value preoperatively has been conclusively demonstrated in patients with osteosarcoma of the extremities. Patients with osteosarcoma of the jaw have a better prognosis with surgery alone, but those with osteosarcoma arising elsewhere in the skull have a poor prognosis and should be treated with chemotherapy as initial therapy. There are four active agents: doxorubicin, ifosfamide (as for soft tissue sarcomas), cisplatin, and high-dose methotrexate. Patients with osteosarcoma of the extremities who show a complete response to preoperative chemotherapy with tumor destruction of at least 90% have significantly improved survival. Insufficient data exist about skull base osteosarcomas, but they are treated in similar fashion. Although a number of regimens have been used, we prefer the combination of doxorubicin 75 mg/m2 IV by 72-hour continuous infusion through a central venous catheter, and cisplatin 120 mg/m2 IV over 4 hours on day 1. If there is more than 90% necrosis, continue the same regimen until cisplatin neurotoxicity and then substitute ifosfamide. If there is less than 90% tumor necrosis at surgery, we would switch to an alternating regimen of high-dose methotrexate, high-dose ifosfamide, and doxorubicin/ifosfamide.
Chondrosarcoma There is no effective chemotherapy regimen for conventional chondrosarcoma of bone. Surgery remains the optimal treatment modality. Investigators at the Massachusetts General Hospital in Boston have championed the use of proton beam radiation for these tumors. The dose can be focused on the target, while achieving significant sparing of the brain, brain stem, cervical cord, and optic nerves and chiasm. For skull base chondrosarcomas, 10-year local control rates with combined proton–photon therapy are 94% (87).
Chordoma Chordomas are rare tumors of bone that originate from the embryonic remnants of the notochord, which is the tissue of derivation for the nucleus pulposus of the intervertebral disks in normal humans. The median age of presentation is around 60 years, with skull base tumors affecting a younger age population. They are typically slow growing tumors, however, they can be locally aggressive and invasive. They most often occur in the axial skeleton and typically arise from the spheno-occipital region of the skull base or the sacrum
Chapter 33: Sarcomas of the Skull Base
(88–93). In adults, chordomas involve the sacrococcygeal region approximately 50% of the time, the base of the skull at the spheno-occipital region approximately 35% of the time, and 15% are found in the vertebral column (88). Craniocervical chordomas often involve the dorsum sella, clivus, and nasopharynx, and therefore, despite the tumors’ propensity to not initially metastasize, they can be extremely locally destructive and debilitating. Three distinct subtypes of chordoma are described: conventional, chondroid, and dedifferentiated. Conventional is the most commonly encountered subtype, and are identified by a lack of cartilaginous or mixed mesenchymal cell types. Chondroid chordomas contain both chordomatous and chondromatous features and can account for up to 33% of cranial chordomas (88). They occur in a younger population compared to conventional chordomas, and generally are less aggressive with a longer median survival. Dedifferentiated chordomas are those that have had a sarcomatous transformation, and make up 2% to 8% of all chordomas (94,95). They can develop at the onset of disease, or later, as the chordoma transforms. Most aggressive chordomas are aneuploid on DNA analysis, as compared to 27% of conventional chordomas, identified in a small DNA flow cytometric study (95). Surgery, with wide margins, can improve overall outcome (96,97). Given the anatomic constraints of chordomas with close proximity to vital structures, this is often impossible, and radiation is commonly used as adjunctive therapy. Utilizing radiation therapy in the management of chordomas is not without its inherent challenges. The tolerance dose of tissues like the brainstem, optic pathway, and spinal cord is much lower than the dose required for cure, approximately 70 Gy (98). Charged particle irradiation, like proton-beam irradiation, has typically been used in an adjuvant manner, enabling the delivery of higher radiation doses to the tumor, while limiting the dose to the surrounding structures, i.e., the eyes or spinal cord (96,99–101). This form of therapy has led to significant improvement in local control for patients with intracranial chordomas. One relatively large study treated 195 patients with chordomas of the base of the skull or cervical spine. At a median follow-up time of 54 months, 69% of patients were relapse free; 5-year and 10-year progression free survival rates of 70% and 45% respectively, were reported (101). There are few studies reporting the use of chemotherapy in chordomas. Although there have been anecdotal responses reported to anthracyclines, cisplatin, and alkylating agents, chemotherapy is generally thought to be inactive, with the exception of dedifferentiated chordomas, which behave more similarly to sarcomas with a propensity to be more responsive to cytotoxic therapy (102). A phase II study of 9-nitro-camptothecin in patients with advanced chordoma or soft tissue sarcoma showed modest activity in delaying progression in patients with unresectable or metastatic chordoma, whereas it had little activity in other soft tissue sarcomas and gastrointestinal stromal tumors (103). Recently, there have been a few small studies evaluating imatinib in the treatment of chordoma. Six cases were described in 2004, with a follow-up compassionate series reported shortly thereafter (104,105). Most patients responded to 800 mg daily of imatinib. Typically, responses were marked by hypodensity and decreased contrast uptake on CT scans. A tumor volume decrease was generally not encountered; however, despite lack of dimensional reduction, there was symptomatic improvement indicating response to therapy. The length of tumor response in a very advanced tumor population was generally 1 year (104,105).
477
Hemangiopericytoma Hemangiopericytoma (HPC) is a rare vascular tumor derived from the pericytes that can be found surrounding all capillaries (also referred to as pericytes of Zimmerman) and are thought to be immature smooth muscle cells of mesenchymal origin (106,107). Pericytes are thought to add to the structural support of blood vessels and play an active role in blood flow regulation. By virtue of their tissue of origin, they can be found virtually anywhere within the body, however, 15% to 25% present in the head and neck, especially associated with the dura. They can have an aggressive clinical course, therefore early diagnosis and implementation of local therapy and in some cases systemic therapy are important in effective management (108,109). The initial management for HPC is surgical resection with attempts at negative margins, if possible. This is oftentimes difficult in tumors situated in the head and neck. In a study published by Soyeur et al., 5-year local control rates after surgery were 84% for patients with gross total resections and 38% for partial resections (110). Postoperative radiation therapy has been shown to play a role in effective local control (111,112). In a study of 37 patients with HPC at a single institution, patients were treated with high-precision radiotherapy or intensity-modulated radiotherapy. Overall survival rates were 100% and 64% at 5 and 10 years, respectively (112). The role of chemotherapy in HPC is less certain. Various chemotherapeutic agents have been used with variable success rates, and given the overall rarity and heterogeneity of tumors, it has been difficult to perform adequate clinical trials. Cyclophosphamide, vincristine, methotrexate, dacarbazine, ifosfamide, and doxorubicin have all been used (113–114) in case series, with doxorubicin generally believed to be the most effective agent. These reports seem overly positive, and one wonders if cases of synovial sarcoma or myxoid liposarcoma, tumors with a prominent hemangiopercytic pattern, may have been included by pathologists with insufficient expertise in sarcomas. One case series utilized interferon alpha, given its antiangiogenesis properties. Tumor responses were seen and continuous freedom from disease progression of 18 to 24 months was reported (115). Other, more targeted antiangiogenic agents, like bevacizumab are currently being evaluated, with data anticipated. Our current approach utilizes a combination of temozolomide and bevacizumab. A preliminary report submitted to ASCO shows benefit in about 80% of patients in this small, 14 patient series (116).
REFERENCES 1. Farhood AI, Hajdu SI, Shiu MH, et al. Soft tissue sarcomas of the head and neck in adults. Am J Surg. 1990;160:365–369. 2. Weber RS, Benjamin RS, Peters LJ, et al. Soft tissue sarcomas of the head and neck in adolescents and adults. Am J Surg. 1986;152:386–392. 3. Freedman AM, Reiman HM, Woods JE. Soft-tissue sarcomas of the head and neck. Am J Surg. 1989;158:367–372. 4. Birkmeyer JD, Siewers AE, Finlayson EV, et al. Hospital volume and surgical mortality in the United States. N Engl J Med. 2002;346:1128–1137. 5. Gustafson P, Dreinhofer KE, Rydholm A. Soft tissue sarcoma should be treated at a tumor center: A comparison of quality of surgery in 375 patients. Acta Orthop Scand. 1994;65:47–50. 6. McKenna WG, Barnes MM, Kinsella TJ, et al. Combined modality treatment of adult soft tissue sarcomas of the head and neck. Int J Radiat Oncol Biol Phys. 1987;13:1127–1133. 7. Figueiredo MT, Marques LA, Campos Filho N. Soft-tissue sarcomas of the head and neck in adults and children: Experience
478
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19. 20.
21.
22.
23.
24.
25.
26.
Thornton and Benjamin at a single institution with a review of literature. Int J Cancer. 1988;41:198–200. Tran LM, Mark R, Meier R, et al. Sarcomas of the head and neck: Prognostic factors and treatment strategies. Cancer. 1992;70:169–177. Le Vay J, O’Sullivan B, Catton C, et al. An assessment of prognostic factors in soft-tissue sarcoma of the head and neck. Arch Otolaryngol Head Neck Surg. 1994;120:981–986. Gil Z, Patel SG, Singh B, et al. For the International Collaborative Study Group. Analysis of prognostic factors in 146 patients with anterior skull base sarcomas: An international collaborative study. Cancer. 2007;110:1033–1041. Prabhu SS, Diaz E, Sturgis EM, et al. Primary sarcomas of the skull base: An analysis of 63 cases. Clin Neurosurg. 2004;51:340– 342. Gerrand CH, Wunder JS, Kandel RA, et al. Classification of positive margins after resection of soft-tissue sarcoma of the limb predicts the risk of local recurrence. J Bone Joint Surg Br. 2001;83:1149. O’Sullivan B, Gullane P, Irish J, et al. Preoperative radiotherapy for adult head and neck soft tissue sarcoma (STS): Assessment of wound complication rates and cancer outcome in a prospective series. World J Surg. 2003;27:875–883. O’Sullivan B, Davis AM, Turcotte R, et al. Preoperative versus postoperative radiotherapy in soft-tissue sarcoma of the limbs: A randomised trial. Lancet. 2002;359:2235–2241. Demetri GD, Elias AD. Results of single-agent and combination chemotherapy for advanced soft tissue sarcomas: Implications for decision making in the clinic. Hematol Oncol Clin North Am. 1995;9:765–785. Bramwell VH, Anderson D, Charette ML. Doxorubicin-based chemotherapy for the palliative treatment of adult patients with locally advanced or metastatic soft-tissue sarcoma: A metaanalysis and clinical practice guideline. Sarcoma. 2000;4:103– 112. Van Glabbeke M, van Oosterom AT, Oosterhuis JW, et al. Prognostic factors for the outcome of chemotherapy in advanced soft tissue sarcoma: An analysis of 2185 patients treated with anthracycline-containing first-line regimens: A European Organization for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group study. J Clin Oncol. 1999;17:150– 157. Benjamin R, Wiernik P, et al. Adriamycin: A new effective agent in the therapy of disseminated sarcomas. Med Pediatr Oncol. 1975;1:63–76. O’Bryan R, Baker L, et al. Dose response evaluation of adriamycin in human neoplasia. Cancer. 1977;39:1940–1948. Mouridsen HT, Bastholt L, Somers R, et al. Adriamycin versus epirubicin in advanced soft tissue sarcomas: A randomized phase II/phase III study of the EORTC Soft Tissue and Bone Sarcoma Group. Eur J Cancer Clin Oncol. 1987;23:1477– 1483. Lopez M, Vici P, Lauro L, et al. Increasing single epirubicin doses in advanced soft tissue sarcomas. J Clin Oncol. 2002;20:1329– 1334. Nielsen OS, Dombernowsky P, Mouridsen H, et al. High-dose epirubicin is not an alternative to standard-dose doxorubicin in the treatment of advanced soft tissue sarcomas: A study of the EORTC Soft Tissue and Bone Sarcoma Group. Br J Cancer. 1998;78:1634–1639. Casper ES, Schwartz GK, Sugarman A, et al. Phase I trial of doseintense liposome-encapsulated doxorubicin in patients with advanced sarcoma. J Clin Oncol. 1997;15:2111–2117. Garcia A, Kempf R, Rogers M, et al. A phase II study of Doxil (liposomal doxorubicin): Lack of activity in poor prognosis soft tissue sarcomas. Ann Oncol. 1998;10:1131–1133. Toma S, Tucci A, Villani G, et al. Liposomal doxorubicin (Caelyx) in advanced pretreated soft tissue sarcomas: A phase II study of the Italian Sarcoma Group (ISG). Anticancer Res. 2000;20:485– 491. Skubitz KM. Phase II trial of pegylated-liposomal doxorubicin (Doxil) in sarcoma. Cancer Invest. 2003;21:167–176.
27. Chidiac T, Budd GT, Pelley R, et al. Phase II trial of liposomal doxorubicin (Doxil) in advanced soft tissue sarcomas. Invest New Drugs. 2000;18:253–259. 28. Antman KH, Montella D, Rosenbaum C, et al. Phase II trial of ifosfamide with mesna in previously treated metastatic sarcoma. Cancer Treat Rep. 1985;69:499–504. 29. Stuart-Harris R, Harper PG, Kaye SB, et al. High-dose ifosfamide by infusion with mesna in advanced soft tissue sarcoma. Cancer Treat Rev. 1983;10(Suppl A):163–164. 30. Benjamin RS, Legha SS, Patel SR, et al. Single-agent ifosfamide studies in sarcomas of soft tissue and bone: The M. D. Anderson experience. Cancer Chemother Pharmacol. 1993;31(Suppl 2):S174–S179. 31. Bramwell VH, Mouridsen HT, Santoro A, et al. Cyclophosphamide versus ifosfamide: Final report of a randomized phase II trial in adult soft tissue sarcomas. Eur J Cancer Clin Oncol. 1987;23:311–321. 32. Elias AD, Eder JP, Shea T, et al. High-dose ifosfamide with mesna uroprotection: A phase I study. J Clin Oncol. 1990;8:170–178. 33. Cinat G, Mickiewicz E, Cabalar M, et al. High dose ifosfamide (HDIFO): A regimen that deserves to be taken into account in previously treated sarcoma bearing patients (Pts). Proc Am Soc Clin Oncol. 2000;19:576a. 34. Deligny N, Demaille M, Bui NB, et al. Phase II trial of high dose of ifosfamide (HDI) in adult patients with advanced soft tissue sarcomas (ASTS). Proc Am Soc Clin Oncol. 2001;20:292b. 35. Le Cesne A, Antoine E, Spielmann M, et al. High-dose ifosfamide: Circumvention of resistance to standard-dose ifosfamide in advanced soft tissue sarcomas. J Clin Oncol. 1995;13:1600–1608. 36. Patel SR, Vadhan-Raj S, Papadopoulos NJ, et al. High-dose ifosfamide in bone and soft tissue sarcomas: Results of phase II and pilot studies—dose-response and schedule dependence. J Clin Oncol. 1997;15:2378–2384. 37. Bui NB, Demaille M, Chevreau C, et al. qMAID vs MAID + 25% with G-CSF in adults with advanced soft tissue sarcoma (STS). First results of a randomized study of the FNCLCC Sarcoma Group. Proc Am Soc Clin Oncol. 1998;17:517a. 38. Frustaci S, Comandone A, Bearz A, et al. Efficacy and tolerability of an ifosfamide continuous infusion (IFO-c.i.) soft tissue sarcoma (STS) patients (pts). Proc Am Soc Clin Oncol. 1998;17:518a. 39. Cerny T, Leyvraz S, von Briel T, et al. Saturable metabolism of continuous high-dose ifosfamide with mesna and GMCSF: A pharmacokinetic study in advanced sarcoma patients. Swiss Group for Clinical Cancer Research (SAKK). Ann Oncol. 1999;10:1087–1094. 40. Gottlieb JA, Baker LH, O’Bryan RM, et al. Adriamycin (NSC123,127) used alone and in combination for soft tissue and bony sarcoma. Cancer Chemother Rep. 1975;6:271–280. 41. Crawford SM, Jerwood D. An assessment of the relative importance of the components of CYVADIC in the treatment of soft-tissue sarcomas using regression meta-analysis. Med Inform (Lond). 1994;19:311–321. 42. Verma S, Bramwell V. Dose-intensive chemotherapy in advanced adult soft tissue sarcoma. Expert Rev Anticancer Ther. 2002;2:201–215. 43. Edmonson JH, Ryan LM, Blum RH, et al. Randomized comparison of doxorubicin alone versus ifosfamide plus doxorubicin or mitomycin, doxorubicin, and cisplatin against advanced soft tissue sarcomas. J Clin Oncol. 1993;11:1269–1275. 44. Santoro A, Tursz T, Mouridsen HT, et al. Doxorubicin versus CYVADIC versus doxorubicin plus ifosfamide in first-line treatment of advanced soft tissue sarcomas: A randomized study of the European Organization for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group. J Clin Oncol. 1995;13:1537–1545. 45. Antman K, Crowley J, Balcerzak SP, et al. An Intergroup phase III randomized study of doxorubicin and dacarbazine with or without ifosfamide and mesna in advanced soft tissue and bone sarcomas. J Clin Oncol. 1993;11:1276–1285. 46. Patel S, Vadhan-Raj S, et al. Results of two consecutive trials of dose-intensive chemotherapy with doxorubicin and
Chapter 33: Sarcomas of the Skull Base
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
ifosfamide in patients with sarcomas. Am J Clin Oncol. 1998;21:317–321. Merimsky O, Meller I, Flusser G, et al. Gemcitabine in soft tissue or bone sarcoma resistant to standard chemotherapy: A phase II study. Cancer Chemother Pharmacol. 2000;45:177–181. Svancarova L, Blay JY, Judson IR, et al. Gemcitabine in advanced adult soft-tissue sarcomas. A phase II study of the EORTC Soft Tissue and Bone Sarcoma Group. Eur J Cancer. 2002;38:556–559. Okuno S, Ryan LM, Edmonson JH, et al. Phase II trial of gemcitabine in patients with advanced sarcomas (E1797): A trial of the Eastern Cooperative Oncology Group. Cancer. 2003;97:1969– 1973. Patel SR, Gandhi V, Jenkins J, et al. Phase II clinical investigation of gemcitabine in advanced soft tissue sarcomas and window evaluation of dose rate on gemcitabine triphosphate accumulation. J Clin Oncol. 2001;19:3483–3489. Talbot SM, Keohan, ML, Hesdorffer, et al. A phase II trial of temozolomide in patients with unresectable or metastatic soft tissue sarcoma. Cancer. 2003;98:1942–1946. Woll PJ, Judson I, Lee SM, et al. Temozolomide in adult patients with advanced soft tissue sarcoma: A phase II study of the EORTC Soft Tissue and Bone Sarcoma Group. Eur J Cancer. 1999;35:410–412. Trent JC, Beach J, Burgess MA, et al. A two-arm phase II study of temozolomide in patients with advance gastrointestinal stromal tumors and other soft tissue sarcomas. Cancer. 2003;98:2693– 2699. Blay JY, Judson I, Rodenhuis S, et al. Phase II study of raltitrexed (Tomudex) for patients with advanced soft tissue sarcomas refractory to doxorubicin-containing regimens. Anticancer Drugs. 1999;10:873–877. De Angelo D, Naujoks R, Manola JB, et al. Phase II study of irinotecan (CPT-11) in relapsed or refractory soft tissue sarcoma (STS). Proc Am Soc Clin Oncol. 2000;19:555a. Carson E, Zalupski M, Redman B, et al. Phase II trial of sargramostim (yeast-derived recombinant human GM-CSF) as monotherapy for advanced sarcomas. Proc Am Soc Clin Oncol. 2000;19:563a. Bramwell VH, Eisenhauer EA, Blackstein M, et al. Phase II study of topotecan (NSC 609 699) in patients with recurrent or metastatic soft tissue sarcoma. Ann Oncol. 1995;6:847–849. Budd GT, Rankin C, Hutchins LF, et al. Phase II trial of topotecan by continuous infusion in patients with advanced soft tissue sarcomas, a SWOG study. Southwest Oncology Group. Invest New Drugs. 2002;20:129–132. Fidias P, Demetri GD, Harmon DC. Navelbine shows activity in previously treated sarcoma patients: Phase II results from MGH/Dana-Farber/Partners Cancer Care Study. Proc Am Soc Clin Oncol. 1998;17:1977. Verweij J, Lee SM, Ruka W, et al. Randomized phase II study of docetaxel versus doxorubicin in first- and second-line chemotherapy for locally advanced or metastatic soft tissue sarcomas in adults: A study of the European Organization for Research and Treatment of Cancer soft tissue and bone sarcoma group. J Clin Oncol. 2000;18:2081–2086. Kostler WJ, Brodowicz T, Attems Y, et al. Docetaxel as rescue medication in anthracycline- and ifosfamide-resistant locally advanced or metastatic soft tissue sarcoma: Results of a phase II trial. Ann Oncol. 2001;12:1281–1288. Edmonson JH, Ebbert LP, Nascimento AG, et al. Phase II study of docetaxel in advanced soft tissue sarcomas. Am J Clin Oncol. 1996;19:574–576. Gian VG, Johnson TJ, Marsh RW, et al. A phase II trial of paclitaxel in the treatment of recurrent or metastatic soft tissue sarcomas or bone sarcomas. J Exp Ther Oncol. 1996;1:186–190. Casper ES, Waltzman RJ, Schwartz GK, et al. Phase II trial of paclitaxel in patients with soft-tissue sarcoma. Cancer Invest. 1998;16:442–446. Delaloge S, Yovine A, Taamma A, et al. Ecteinascidin-743: A marine-derived compound in advanced, pretreated sarcoma patients—preliminary evidence of activity. J Clin Oncol. 2001;19:1248–1255.
479
66. Taamma A, Misset, JL, Riofrio M, et al. Phase I and pharmacokinetic study of ecteinascidin-743, a new marine compound, administered as a 24-hour continuous infusion in patients with solid tumors. J Clin Oncol. 2001;19:1256–1265. 67. George S, Maki RG, Harmon DC, et al. Phase II study of ecteinascidin-743 (ET-743) given by 3-hour IV infusion in patients (pts) with soft tissue sarcomas (STS) failing prior chemotherapies. Proc Am Soc Clin Oncol. 2002;21:408a. 68. Le Cesne A, Blay J-Y, Judson I, et al. Phase II study of ET-743 in advanced soft tissue sarcomas: A European Organisation for the Research and Treatment of Cancer (EORTC) soft tissue and bone sarcoma group trial. J Clin Oncol. 2005;23:576–584. Erratum in: J Clin Oncol. 2005;23:5276. 69. Dileo P, Casali P, Bacci G, et al. Phase II evaluation of 3-hour infusion ET-743 in patients with recurrent sarcomas. Proc Am Soc Clin Oncol. 2002;21:408a. 70. Grosso F, Jones RL, Demetri GD, et al. Efficacy of trabectedin (ecteinascidin-743) in advanced pretreated myxoid liposarcomas: A retrospective study. Lancet Oncol. 2007;8:595– 602. 71. Skubitz KM, Hamdan H, Thompson RC Jr. Ambulatory continuous infusion ifosfamide with oral etoposide in advanced sarcomas. Cancer. 1993;72:2963–2969. 72. Yalcin S, Gullu I, Barista I, et al. Treatment of advanced refractory sarcomas with ifosfamide and etoposide combination chemotherapy. Cancer Invest. 1998;16:297–302. 73. Saeter G, Alvegard TA, Monge OR, et al. Ifosfamide and continuous infusion etoposide in advanced adult soft tissue sarcoma. A Scandinavian Sarcoma Group Phase II Study. Eur J Cancer. 1997;33:1551–1558. 74. Papai Z, Bodoky G, Szanto J, et al. The efficacy of a combination of etoposide, ifosfamide, and cisplatin in the treatment of patients with soft tissue sarcoma. Cancer. 2000;89:177– 180. 75. Hensley ML, Maki R, Venkatraman E, et al. Gemcitabine and docetaxel in patients with unresectable leiomyosarcoma: Results of a phase II trial. J Clin Oncol. 2002;20:2824–2831. 76. Leu K, Ostruszka LJ, Shewach D, et al. Laboratory and clinical evidence of synergistic cytotoxicity of sequential treatment with gemcitabine followed by docetaxel in the treatment of sarcoma. J Clin Oncol. 2004;22:1706–1712. 77. Thall P. Wathen J. Covariate-adjusted adaptive randomization in a sarcoma trial with multi-stage treatments. Stat Med. 2005;24:1947–1964. 78. Maki R, Wathen J, Patel SR, et al. Randomized phase II study of gemcitabine and docetaxel compared with gemcitabine alone in patients with metastatic soft tissue sarcomas: Results of sarcoma alliance for research through collaboration study 002 [corrected]. J Clin Oncol. 2007;25:2755–2763. 79. Rosen G, Forscher C, Lowenbraun S, et al. Synovial sarcoma: Uniform response of metastases to high dose ifosfamide. Cancer. 1994;73:2506–2511. 80. Kampe CE, Rosen G, Eilber F, et al. Synovial sarcoma: A study of intensive chemotherapy in 14 patients with localized disease. Cancer. 1993;72:2161–2169. 81. Nielsen OS, Judson I, van Hoesel Q, et al. Effect of high-dose ifosfamide in advanced soft tissue sarcomas. A multicentre phase II study of the EORTC Soft Tissue and Bone Sarcoma Group. Eur J Cancer. 2000;36:61–67. 82. Edmonson JH, Ryan L, Blum RH. Phase II study of ifosfamide + doxorubicin in patients with advanced synovial sarcomas: (E1793): A Trial of the Eastern Cooperative Oncology Group. Sarcoma. 2003;7:9–11. 83. Demetri GD, Spiegelman BM, Fletcher CDM, et al. Differentiation of liposarcomas in patients treated with the PPAR-g ligand troglitazone: Documentation of biologic activity in myxoid/round cell and pleomorphic subtypes. Proc Am Soc Clin Oncol. 1999;18:535a. 84. Demetri GD, Fletcher CD, Mueller E, et al. Induction of solid tumor differentiation by the peroxisome proliferator-activated receptor-gamma ligand troglitazone in patients with liposarcoma. Proc Natl Acad Sci USA. 1999;96:3951–3956.
480
Thornton and Benjamin
85. Ferrari A, Dileo P, Casanova M, et al. Rhabdomyosarcoma in adults: A retrospective analysis of 171 patients treated at a single institution. Cancer. 2003;98:571–580. 86. Little DJ, Ballo MT, Zagars GK, et al. Adult rhabdomyosarcoma: Outcome following multimodality treatment. Cancer. 2002;95:377–388. 87. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull base. Strahlenther Onkol. 1999;175(Suppl 2):57–63. 88. Heffelfinger MJ, Dahlin DC, MacCarty CS, et al Chordomas and cartilaginous tumors at the skull base. Cancer. 1973;32:410–420. 89. Perzin KH and Pushparaj N. Nonepithelial tumors of the nasal cavity, paranasal sinuses and nasopharynx. A clinicopathological study. XIV: Chordomas. Cancer. 1986;57:784–796. 90. Harbour JW, Lawton MT, Criscuolo GR, et al. Clivus chordoma: A report of 12 recent cases and review of the literature. Skull base Surg. 1991;1:200–206. 91. Volpe R, Mazabraud A. A clinicopathologic review of 25 cases of chordoma (a pleomorphic and metastasizing neoplasm). Am J Surg Path. 1983;7:161–170. 92. Rich TA, Schiller A, Suit HD, et al. Clinical and pathologic review of 48 cases of chordoma. Cancer. 1985;56:182–187. 93. Mirra J, et al. Chordoma. In: Fletcher CD, Unni K, Mertens F, eds. World Health Organization of Tumours: Pathology and Genetics of Tumours of Soft Tissue and Bone. Lyon, France: IARC Press, 2002:316–317. 94. Meis JM, Raymond AK, Evans HL, et al. Dedifferentiated chordoma. A clinicopathologic and immunohistochemical study of three cases. AM J Surg Path. 1987;11:516–525. 95. Hruban RH, Traganos F, Reuter VE, et al. Chordomas with malignant spindle cell components. A DNA flow cytometric and immunohistochemical study with histogenetic implications. Am J Pathology. 1990;137:435–447. 96. Carpentier A, Polivka M, Blanquet A, et al. Suboccipital and cervical chordomas: The value of aggressive treatment at first presentation of disease. J Neurosurge. 2002;97:1070–1077. 97. Tzortzidis F, Elahi F, Wright D, et al. Patient outcome at longterm follow-up after aggressive microsurgical resection of cranial based chordomas. Neurosurg Online. 2006;59:230–237. 98. Casali PG, Stacchiotti S, Sangalli C, et al. Chordoma. Curr Opin Oncol. 2007;19(4):367–370. 99. Hug EB, Loredo LN, Slater JD, et al. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. J Neurosurg. 1999;91:432–439. 100. Fagundes MA, Hug EB, Liebsch NJ, et al. Radiation therapy for chordomas of the base of the skull and cervical spine: Patterns of failure and outcome after relapse. Int J Radiat Oncol Biol Phys. 1995;33:579–584.
101. Debus J, Hug EB, Liebsch NJ, et al. Brainstem tolerance to conformal radiotherapy of skull base tumors. Int J Radiat Oncol Biol Phys. 1997;39:967–975. 102. Fleming GF, Heimann PS, Stephens JK, et al. Dedifferentiated chordoma. Response to aggressive chemotherapy in two cases. Cancer. 1993;72:714–718. 103. Chugh R, Dunn R, Zalupski MM, et al. Phase II study of 9-nitrocamptothecin in patients with advanced chordoma or soft-tissue sarcoma. JCO. 2005;23;3597–3604. 104. Casali PG, Messina A, Stacchiotti S, et al. Imatinib mesylate in chordoma. Cancer. 2004;101;2086–2097. 105. Casali PG. Imatinib mesylate in 18 advanced chordoma patients [abstract]. J Clin Onc. 2005;23(suppl):9012. 106. Stout AP, Murray MR. Hemangiopericytoma: A vascular tumor featuring Zimmerman’s pericytes. Ann Surg. 1942;116:26– 33. 107. Stout AP. Hemangiopericytoma. A study of 25 new cases. Cancer. 1949;2:1027–1035. 108. Batsakis JG, Rice DH. The pathology of head and neck tumors: Vasoformative tumors, part 9B. Head Neck Surg. 1981;3:326– 340. 109. McMaster MJ, Soule EH, Ivins JC. Hemangiopericytoma. A clinicopathologic study and long-term follow-up of 60 patients. Cancer. 1975;36:2232–2244. 110. Soyuer S, Chang EL, Selek U, et al. Intracranial meningeal hemangiopericytoma: The role of radiotherapy: Report of 29 cases and review of the literature. Cancer. 2004;100;1491–1497. 111. Dufour H, M´etellus P, Fuentes S, et al. Meningeal hemangiopericytoma: A retrospective study of 21 patients with special review of postoperative external radiotherapy. Neurosurgery. 2001;48:756–762. 112. Combs S, et al. Precision radiotherapy for hemangiopericytomas of the central nervous system. Cancer. 2005;104:2457– 2465. 113. Beadle GF, Hillcoat BL. Treatment of advanced malignant hemangiopericytoma with combination adriamycin and DTIC: A report of four cases. J Surg Onc. 1983;22;167–170. 114. Wong PP, Yagoda A. Chemotherapy of malignant hemangiopericytoma. Cancer. 1978;41:1256–1260. 115. Kirn DH, Kramer A. Long-term freedom from disease progression with interferon alpha therapy in two patients with malignant hemangiopericytoma. J Natl Cancer Inst. 1996;88:764– 765. 116. Park MS, Patel SR, Ludwig JA, et al. Combination therapy with temozolomide and bevacizumab in the treatment of hemangiopericytoma and malignant solitary fibrous tumor. J Clin Oncol. 26: 2008 (May 20 suppl; abstr 10512).
34 Angiofibromas and Vascular Tumors of the Skull Base Andrew G. Sikora and Randal S. Weber
that the majority of putative JNA occurring in nonadolescent males, and sporadic reports of JNA in females, is misdiagnosed. JNA is thought to be more common in Egypt and India, and possibly in other areas of South Asia, as well as in Kenya (3). Of note, this geographic distribution does not correspond to the distribution of areas where nasopharyngeal carcinoma is endemic, and does not seem to provide clues to genetic or environmental factors predisposing to JNA. Unlike JNA, HPC affects males and females with equal frequency and usually presents after the second decade (4,5). Hemangiomas in general have a female-to-male ratio of 2– 4:1 and usually present in the first year of life (6). Vascular malformations have a slight female predominance. The rarity of each of these lesions makes it unclear whether these trends are as true for lesions limited to the skull base as at other body sites.
INTRODUCTION Vascular tumors of the skull base are both rare and diverse. Despite the heterogeneity of lesions presenting in this area, they share common principles of management and surgical approach, including reliance on radiology to establish diagnosis, surgery as the primary treatment, and a vital role for angiography as both diagnostic and adjunctive therapeutic modality. With the exception of the malignant subtype of hemangiopericytoma (HPC), these tumors are generally benign but locally aggressive due to their proximity to vital structures and potential for intracranial spread. The anatomic and functional complexity of this region mandates a collaborative, multidisciplinary approach, which may require the participation of otolaryngology/head and neck surgery, neurosurgery, neuroradiology, interventional radiology, reconstructive surgery, ophthalmology, speech and swallowing therapeutics, and other teams. In the present chapter, we seek to provide an overview of the multidisciplinary management of vascular tumors of the base of skull, written from the skull base surgeon’s perspective. To provide a paradigm for management of these lesions, we describe in detail the evaluation and management of juvenile nasopharyngeal angiofibroma (JNA), the most common vascular tumor presenting in this region. Where appropriate we discuss other tumor types separately including hemangiomas, vascular malformations, HPC, and malignant HPC.
PATHOLOGY Juvenile Nasopharyngeal Angiofibroma Etiology and Pathogenesis Although most patients diagnosed with JNA undergo treatment, its natural history is thought to be androgen-driven growth during puberty followed by regression. Support for the concept that JNA can involute after puberty is provided by several case reports of lesions which underwent radiographically documented regression during observation or after incomplete excision (7,8). Despite the tantalizing clue provided by the nearexclusive occurrence of JNA in adolescent boys, and numerous theories proposed to explain its pathogenesis, the etiology of JNA is still unclear. Earlier histologists believed that JNA was primarily a fibrous tumor, which led to theories that JNA arose from hypertrophy of skull base periosteum or fascia, or inappropriate persistence of fibrocartilage rests. Later, appreciation of the predominantly vascular nature of the tumor, and similarity to erectile tissue, caused researchers to speculate that JNA results from misplaced inferior turbinate tissue, or other vascular tissue. The nature of the lesion— neoplasm or hamartoma—has also been a source of controversy, although a recent study that demonstrated frequent beta-catenin mutations in JNA specimens lends support to the idea that JNA is a clonal, neoplastic process (9). The only clue to genetic predisposition of JNA comes from patients with familial adenomatous polyposis (FAP), who have up to a 25-fold increase in the frequency of JNA (10–13). Since genes in the adenomatous polyposis coli (APC) gene pathway are commonly mutated in FAP, and APC regulates binding and degradation of beta-catenin, this lends additional support to the involvement of beta-catenin in pathogenesis of JNA. Of note, the study (9) which found increased
EPIDEMIOLOGY AND INCIDENCE Incidence Vascular lesions of the skull base are rare; the most common lesion, JNA, makes up less than 0.05% of all head and neck tumors (1). HPC is thought to account for less than 1% of all vascular tumors at all body sites (1). From 9% to 28% of HPCs have been estimated to occur in the head and neck, with sinonasal presentation being most common (1). The rarity of other vascular skull base lesions makes it difficult to estimate their incidence in this location. Paragangliomas are described in detail in a separate chapter of this book.
Epidemiology The epidemiology of JNA has proven fascinating to clinicians and researchers, who seek clues to its etiology in its demographic characteristics. JNA almost exclusively affects adolescent boys, and is seldom observed in young adults—this has led to speculation that the tumor develops in response to the hormonal milieu occurring during male puberty (see below) and regresses upon reaching adulthood. The average age of patients with JNA is 14 to 18 years, although the reported range is much broader (7–29 years) (2). It is suspected 481
482
Sikora and Weber
mutations of beta-catenin in JNA specimens examined tumors from patients with sporadic, not FAP-associated disease. Beta-catenin is a cytoplasmic protein which is involved in cell–cell adhesion, and also plays a role in oncogenic signal transduction—however, while the association between beta-catenin pathway mutations and JNA deserves further study, it does not yet provide a coherent picture of JNA pathogenesis. The peculiar association of JNA and adolescent males leaves little doubt that whatever the genetic/developmental events leading to susceptibility to JNA, its progression is under hormonal control. This has led to substantial interest in the staining of JNA specimens for receptors for androgenic and other hormones. The clinical behavior of JNA is quite consistent with androgen-dependent growth, and androgen receptors are frequently found on JNA specimens, localized to both stromal and endothelial cells (14,15). Staining of other hormonal receptors is less consistent, with positive staining for progesterone receptors observed sometimes, and staining for estrogen receptors observed rarely if at all (16,17). The geographical variation in JNA incidence and occurrence of JNA in some—but not all—FAP patients has led to interest in environmental factors which may contribute to JNA development. Proposed environmental factors, such as environmental toxins, trauma, dry air, infectious agents, allergy, and other causes of inflammation, are supported by little or no published data.
(A)
Anatomy Although established JNA may have broadly based or multiple attachments to the nasal cavity, it is thought to originate on the posterolateral wall of the nasal cavity, where the root of the pterygoid process, horizontal ala of the vomer, and sphenoidal process of the palatine bone meet (superior aspect of the sphenopalatine foramen) (18,19). This location has great significance because it allows JNA to grow into the nasal cavity, nasopharynx, or pterygopalatine fossa (PPF), allowing further spread to vital areas such as the infratemporal fossa, cranial cavity, and orbit. A slightly different origin has been suggested by Lloyd and colleagues, who propose that JNA originates “in the pterygopalatine fossa in the recess behind the sphenopalatine ganglion, at the anterior aperture of the pterygoid canal” based on imaging characteristics in a series of 72 patients (20). Despite the number of anatomical regions potentially accessible to JNA, it tends to spread in an orderly fashion via one of several pathways (Fig. 1). 1. Anteriorly into the nasal cavity, causing nasal obstruction. 2. Posteriorly into the nasopharynx, where it can access the sphenoid sinus, and may displace the palate downward or even protrude into the oral cavity. 3. Laterally via the sphenopalatine foramen or erosion of the posterior maxillary sinus wall into the pterygopalatine fossa; from there it may spread through the pterygomaxillary fissure to involve the infratemporal and even temporal fossae. 4. Superiorly into the orbit via the inferior orbital fissure. JNA can choose any of these pathways, and can spread in multiple directions simultaneously. The pathways of spread involved determine the presenting symptoms, and have great implications for choice of surgical approach. Continued spread can lead to the most challenging management aspects of JNA, intracranial involvement, via one of several well-defined pathways.
(B)
Figure 1 Routes of invasion of angiofibroma. (A) Sagittal view demonstrating spread of angiofibroma anteriorly into the nasal cavity (1), posteriorly into the nasopharynx (2), and intracranially via the sphenoid sinus (3). (B) Axial view, demonstrating spread of angiofibroma anteriorly into the nasal cavity (1), posteriorly into the nasopharynx (2), and laterally (3) toward the infratemporal and temporal fossae (3a).
1. Via the pterygopalatine fossa/infratemporal fossa, by eroding the bone of the anterior face of the greater wing of the sphenoid. Entry occurs through a region demarcated by the foramen rotundum, foramen ovale, and foramen lacerum. 2. Via the superior orbital fissure. 3. Superiorly, through the roof of the sphenoid sinus. Pathways #1 and #2 are lateral pathways, which result in extension of JNA lateral to the carotid artery and cavernous sinus. Pathway #3, the medial pathway, is less common but results in tumor spread to a much less favorable location medial to the carotid and cavernous sinus. This allows tumor to infiltrate the pituitary and optic chiasm, and makes surgical removal much more difficult. Direct spread superiorly through the cribriform plate is possible, but is extremely rare. The blood supply to the tumor is variable, but in most cases the main blood supply originates from the ipsilateral internal maxillary artery (IMA) (21). As the tumor grows, other vessels can become parasitized, including other branches of the external carotid system such as the sphenopalatine and ascending pharyngeal arteries, as well as branches of the vertebral and internal carotid arteries, such
Chapter 34: Angiofibromas and Vascular Tumors of the Skull Base
as the mandibulovidian artery which comes off the internal carotid artery (ICA). In each case, the tumor can recruit vessels from either the ipsilateral or contralateral side, mandating bilateral angiographic evaluation (see below).
Pathology The gross appearance of JNA is of a firm/spongy lesion, which is pinkish where covered by mucosa and gray or whitish where not. While the tissue lacks a true capsule, it usually has a well-defined pseudocapsule and is sharply demarcated from the surrounding tissue (22). Microscopically, JNA is composed of dense fibrous stroma, punctuated by numerous blood vessels of varying size and shape, which may have a “staghorn” appearance or be slit-like (Fig. 2) (22,23). The fibrous stroma is composed of plump/polygonal fibroblasts with round or vesicular nuclei and abundant connective tissue. The vessels are delicate and consist of a single
Table 1
483
Mulliken and Glowacki Classification of Vascular Lesions
Hemangioma Capillary Cavernous Combined
Vascular malformation “Low flow” Venous Capillary Lymphatic
“High flow” Arteriovenous Arterial A-V fistula
Source: Adapted from Ref. 6.
layer of endothelial cells without an elastic or smooth muscle layer—hence the propensity for these lesions toward massive, uncontrollable bleeding (22). Other fibrovascular neoplasms from which JNA must be differentiated include solitary fibrous tumor and HPC. In the case of solitary fibrous tumor, although the vessels may have a similar appearance, the stromal cells will appear more spindle-shaped with elongated nuclei (24,25). In HPC, the stromal cells may appear similar, but the vessels tend to be round and staghorn-shaped vessels are less common (26).
Hemangiopericytoma
(A)
HPCs are thought to arise from the vascular pericytes of Zimmerman, mesenchymal cells which line blood vessels and regulate blood flow and vessel contraction (27). Little is known about their etiology and pathogenesis, and some pathologists question the validity of HPC as a distinct pathologic entity (28). HPC can present as a spectrum of benign, malignant, and borderline phenotypes, although it has been suggested that HPC of the sinonasal area is more likely to be benign than at other sites (29). Lymphatic dissemination rarely occurs, but hematologic metastasis to the lungs, liver, and bone can occur. The natural history of these tumors is variable, with even histologically benign-appearing tumors sometimes becoming locally aggressive and metastasizing (30). Grossly, HPC can have a variable appearance: soft, firm, rubbery, or polypoid, and it can be tan, gray, or offwhite in color. Unlike JNA, HPC is sometimes mistaken for nasal polyps. The histological appearance is also variable, and can feature small, tightly packed cells with sparse cytoplasm and vesicular nuclei, although cells can also be spindle shaped (31). As is the case with JNA, the stroma is punctuated with thin, fragile vessels, which may occasionally take on the staghorn shape associated with that lesion. Immunostaining for smooth muscle actin is often observed and supports a pericystic origin for these tumors (31). Although solitary fibrous tumor is more likely to have a benign course than HPC, it can be difficult to differentiate these tumors (28); CD34 has been used for this purpose because solitary fibrous tumors tend to stain intensely, whereas staining of HPC tends to be weaker and more focal (24).
Hemangiomas and Vascular Malformations (B)
Figure 2 Gross and microscopic histological appearance of angiofibroma. (A) Gross specimen. Note lobulation and color variation from light tan to hemorrhagic appearance. (B) Characteristic histopathological appearance (original magnification 100X) of angiofibroma demonstrating thin, lymphaticlike endothelial channels without muscular or connective tissue layers. Inset is a high power view (original magnification 400X) demonstrating stromal cells with plump, ovoid nuclei. Source: Image courtesy of Dr. Michelle Williams, UT MD Anderson Cancer Center.
While classification of vascular anomalies can be controversial, the system of Mulikin and Glowaki is commonly accepted Table 1 (6). This classification distinguishes between hemangiomas and vascular malformations, which are further subdivided into high-flow (arteriovenous malformation) and low-flow (various lymphatic, venous, and capillary malformations) lesions. These distinctions are important because the natural histories of hemangioma and vascular malformation differ—hemangiomas undergo intense proliferation during the first months/years of life, followed by gradual involution, which may be partial or complete. They rarely involve bone. Vascular malformations develop in utero, and subsequently
484
Sikora and Weber
Table 2 Radkowski Modification of the Sessions Classification of JNA
Table 3
Tumor stage
Classification
IA
Tumor limited to posterior nares and/or nasopharyngeal vault Tumor involving the posterior nares and/or nasopharyngeal vault with involvement of at least 1 paranasal sinus Minimal lateral extension into the pterygomaxillary fossa Full occupation of the pterygomaxillary fossa with or without superior erosion of the orbital bones Extension into the infratemporal fossa or extension posterior to the pterygoid plates Erosion of the base of skull (middle cranial fossa/base of pterygoids)—minimal intracranial extension Extensive intracranial extension with or without extension into the cavernous sinus
A. Sessions IA Tumor limited to nose and/or nasopharyngeal vault IA Extension into 1 or more paranasal sinuses IIA Minimal extension into pterygomaxillary fissure IIB Full occupation of pterygomaxillary fissure with or without erosion of orbital bones IIC Extension to infratemporal fossa +/− involvement of cheek III Intracranial extension
IB
IIA IIB IIC IIIA IIIB
Source: Adapted from Ref. 36.
Other Classification Systems
B. Fisch I Tumor limited to nasal cavity and/or nasopharynx with no bone destruction II Tumor invading pterygomaxillary fissure or paranasal sinuses with bony destruction III Tumor invading infratemporal fossa, orbit, and/or parasellar region remaining lateral to cavernous sinus IV Tumor invading cavernous sinus, optic chiasm, and/or pituitary fossa Source: Adapted from Refs. 37 and 82.
grow in proportion to the growth of the child, and do not involute. They are more likely to involve bone. Hemangiomas of the skull base are rare, with the orbital apex being the most common site; lesions in this area can cause loss of vision by compression of the optic nerve or affect other structures passing through the superior orbital fissure. Conversely, true arteriovenous malformations are rarely limited to the orbit, and commonly arise intracranially. The etiology of hemangiomas and vascular malformations is not well understood. Hemangiomas are thought to be neoplasms which undergo a postnatal proliferation in response to abnormally regulated mediators of growth, angiogenesis, or inflammation, such as vascular endothelial growth factor, basic fibroblast growth factor, and transforming growth factor–beta (TGF-β) (32). Although some hemangiomas seem to arise in response to trauma, the majority are congenital. Vascular malformations are thought to be anomalies of vascular development, possibly due to defects in the ability of fetal mesenchyme to form endothelium and supporting structures beginning at 8 weeks of gestation. Most are sporadic, although some syndromes are associated with vascular malformations; Sturge-Weber syndrome (33) and von Hippel Lindau (34) syndrome have the potential for skull base involvement due to the predisposition toward vascular malformations in the distribution of the trigeminal and optic nerve or retina, respectively. Histopathological appearance of hemangiomas is diverse, with capillary, cavernous, and mixed types recognized (6). The vascular wall is similar to that of normal vessels and consists of mature endothelium. Mast cell infiltration may be prominent, especially during the involution phase. In vascular malformations, the vessels are dilated and ectatic, and consist of a single endothelial cell layer surrounded by a markedly attenuated muscular layer (35).
STAGING Numerous staging systems have been developed for JNA Tables 2 and 3. The most recently developed system in wide usage is the modification, by Radkowski and colleagues (36), of the staging system developed by Sessions and colleagues Table 2 (37). The Sessions classification reflects the proclivity of JNA to extend into the pterygopalatine fissure and infratemporal fossa, and it also reflects the prognostic impact of invasion of these areas and intracranial extension.
The Radkowski modification reflects the principle that progressively higher levels of involvement of the skull base lead to a greater chance of recurrence, and that extension posterior to the pterygoid plates into the medial and lateral pterygoid muscles complicates surgical excision. The Radkowski modification also differentiates between patients with skull base erosion or minimal intracranial spread and patients with extensive intracranial involvement, since this has implications for both management and prognosis.
CLINICAL ASPECTS Symptoms Symptoms of vascular skull base lesions are determined primarily by the anatomical location and extent of the lesion. In general, presentation at early stages is due to symptoms of nasal obstruction and episodic epistaxis. Since these symptoms are very nonspecific, particularly in the pediatric age group, diagnosis is often delayed and most patients have been symptomatic for months before referral to an otolaryngologist or skull base surgeon. While epistaxis is most commonly periodic and self-limited, the thin, fragile vessels of JNA lack a muscular layer, and thus cannot vasoconstrict in response to hemorrhage. Thus massive, uncontrollable nosebleeds can occur spontaneously or in response to seemingly minor trauma. Extension of JNA (and other lesions) within the nasopharynx and nasal cavity can lead to hyponasal speech; obstruction of the Eustachian tube with otalgia, effusion, and conductive hearing loss; mouth breathing, snoring, and obstructive sleep apnea; and contralateral nasal obstruction from bowing of the nasal septum to the opposite side. Prolonged obstruction can lead to anosmia, mucopurulent nasal discharge, and obstructive sinusitis. As these lesions progress, they gradually erode the skull base leading to invasion of critical regions such as the pterygopalatine and infratemporal fossae, the orbit, the optic chiasm, and sella turcica, leading to upper cranial nerve deficits and threatening vision (see description of progressive spread of JNA above). In this case, symptoms correspond to the area of invasion. Invasion of the orbit can lead to exophthalmos, visual compromise via compression of the optic nerve, or diplopia, which may be due to direct muscle impingement or involvement of the extraocular muscles. Characteristic deficits are produced by involvement of the superior
Chapter 34: Angiofibromas and Vascular Tumors of the Skull Base
485
tion to perform a transnasal biopsy in the office, since severe bleeding may result. The diagnosis of these lesions is usually suggested by the clinical history and physical examination, and is confirmed radiographically.
Radiology
Figure 3 Typical appearance of adolescent male with JNA, which has spread laterally to involve the cheek.
orbital fissure (III, IV, V1, and VI); these deficits plus involvement of the optic nerve suggest spread to the orbital apex, and deficit of V2 suggests involvement of the cavernous sinus (38). Involvement of the pterygopalatine and infratemporal fossae may produce midfacial/dental anesthesia, headache, and excessive lacrimation. Involvement of the temporal fossa can result in trismus, or difficulty in chewing. Spread of tumor outside the confines of the skull base can lead to a visible or palpable swelling of the cheek, temple, or intraorally (Fig. 3). It should be emphasized that considerable infiltration of these anatomic regions can occur before becoming symptomatic, so clinicians must have a high index of suspicion when evaluating patients with nasal obstruction, epistaxis, and cranial nerve deficits. Continued erosion of the skull base leads to intracranial extension, with the potential for symptoms related to involvement of the cavernous sinus (see above), dura, and brain parenchyma including headache, seizures, central neurological deficits, and cranial nerve deficits caused by traction and tenting of the nerve by tumor.
Physical Examination A thorough head and neck evaluation, including meticulous documentation of cranial nerve status and nasopharyngoscopy, should be performed. Splaying of the nasal bones or facial contour deformity may be observed in long-standing cases. Anterior rhinoscopy may reveal bowing of the septum to the opposite side, nasal obstruction, and nasal discharge, which may be mucoid or mucopurulent. Examination of the posterior nasal cavity and nasopharynx with a rigid or flexible scope may display a pink or reddish mass filling these areas. If a vascular mass is suspected (e.g., in an adolescent male with nasal obstruction), one should avoid the tempta-
Imaging is of vital importance in the workup of vascular skull base lesions to confirm the suspected diagnosis, to determine the extent of disease and structures involved, and to serve as a road map for surgical planning (Figs. 4 and 5). CT is the cornerstone of imaging these lesions because it provides excellent resolution of the structures involved and extent of bony invasion. MRI is useful for evaluating extension to soft tissue, and particularly for evaluating cavernous sinus involvement and the extent of intracranial disease. MRI is also excellent for distinguishing invasion of paranasal sinuses by tumor from postobstructive opacification. The anatomic origin of JNA at the sphenopalatine foramen and pterygopalatine fossa leads to two consistent diagnostic features on axial CT scan: a mass involving the nasal cavity and pterygopalatine fossa, and bony erosion behind the sphenopalatine foramen at the root of the medial pterygoid plate (38). These features were seen in every patient in the series of Lloyd and colleagues (38); other areas of the skull base (e.g., nasopharynx, infratemporal fossa) are sometimes, but not always, involved. Another consistent feature is expansion of the pterygoid (vidian) canal, associated with bony erosion of the pterygoid process. Other findings include expansion of the orbital fissure and erosion of the maxillary sinus or basisphenoid. The antral bowing sign described by Holman and Miller (originally described for plain skull films, but better seen on axial CT), in which the posterior wall of the maxillary antrum is displaced forward, can be seen in slow-growing lesions other than angiofibroma, and is not pathognomonic. On CT, angiofibromas avidly enhance after contrast bolus. On MRI, the lesion may be hypo- to isointense on T1 sequences, hyperintense on T2, and enhances on T1 sequence with gadolinium contrast; signal voids consistent with hypervascularity may be seen (39). Imaging plays a vital role in the staging and preoperative planning of JNA. It is important to identify findings which predict a greater risk of recurrence and more difficult surgical resection, including involvement of the anterior fossa, sphenoid, pterygoid muscles or base of the pterygoid plates, or foramen lacerum (39). After treatment, imaging is used to enhance surveillance for recurrence, since tumor regrowth can be quite extensive before becoming clinically detectable. Imaging is also important in the evaluation of HPC, hemangioma and vascular malformations, to confirm the extent of disease, to allow planning for surgery, and to follow for recurrence. These lesions can have similar imaging characteristics (40). They tend to be isointense to muscle on CT with avid contrast enhancement. On MRI, they are iso- to hyperintense to muscle on T1-weighted sequences, hyperintense on T2-weighted sequences, and enhance markedly with gadolinium contrast. Hemangiomas often show a serpentine pattern of vascular structure, and may have either infiltrative or well-demarcated margins (41).
Angiography Angiography plays a role in both the diagnosis and the treatment of JNA, as preoperative embolization is widely recommended to reduce intraoperative hemorrhage and has even been proposed as definitive therapy. Angiography confirms the vascular nature of the tumor and allows delineation of
486
Sikora and Weber
A
B
Figure 4 Axial contrast-enhanced CT scan of the paranasal sinuses windowed for soft tissue (A) and (B) bone. The angiofibroma is seen as a contrast-avid mass, which occupies the right posterior nasal vault with minimal lateral extension into the pterygopalatine fossa, and is thus staged as a Radkowski IIA. Note invasion of the right sphenoid bone with widening of the right vidian canal in panel B (arrow). This lesion would be amenable to a maxillotomy approach supplemented with endoscopic visualization, with use of the high-speed drill to extirpate disease from the area of the vidian canal. Source: Images courtesy of Laurence Ginsberg, UT MDACC Cancer Center.
the vascular anatomy and blood supply. Generally, the tumor manifests as an intense and homogenous blush on angiography (Fig. 6). As described above, a branch of the external carotid system (usually the ipsilateral IMA) is the primary blood supply, but as the tumor grows, it can establish connections with ispsilateral or contralateral branches of both the external and internal carotid systems or vertebral arteries (39). Therefore, bilateral examination of the internal and external carotids and vertebral arteries should be performed in all cases. Embolization of feeding vessels has been recommended to reduce intraoperative blood loss and found to be effective in a number of studies (42–44). Embolization is typically performed transarterially, although direct puncture embolization of the lesion has also been described, and may be useful in cases where intravascular access to the tumor is difficult or likely to result in spillage of the emoblization particles into the internal carotid system with threat to the central nervous system or ophthalmic artery (45,46). The optimal time of embolization is thought to be 24 to 48 hours prior to surgery to allow for maximal devascularization without formation of collateral vessels. Some authors question the effectiveness of embolization since not all series show a reduction in intraoperative blood loss (47,48), and it has been proposed that postembolization tumor shrinkage can increase the risk of incomplete resection, especially where there is deep invasion of the sphenoid (20). However, we feel that for larger lesions, and in cases where a difficult or lengthy resection is anticipated, the benefits of embolization outweigh the potential disadvantages. Although embolization has been proposed as single therapeutic modality for selected JNA, most surgeons feel that it is better used in combination with definitive surgery. Balloon occlusion testing should be considered for patients with lesions involving the cavernous sinus where sacrifice of or injury to an internal carotid artery is a significant possibility. Angiography can also be helpful in the workup and management of other vascular lesions. In the case of hemangioma, angiography allows discrimination and vascu-
lar mapping of high-flow lesions (which may benefit from embolization) from low-flow lesions (which usually do not) (49,50). Angiography can also provide valuable pretreatment information about vascular malformations and HPCs; here too, preoperative embolization may be indicated for larger lesions. For lesions where reconstruction with a local flap (such as the temporalis muscle flap) is required, this fact must be communicated to the interventional radiologist to ensure that vascular compromise is avoided.
Other Considerations The issue of pretreatment biopsy for these vascular lesions is controversial. In the majority of cases, the clinical history and imaging findings provide adequate information for diagnosis and planning, and biopsy can be deferred to the time of definitive surgery. Even in the case of HPC, which has both benign and malignant presentations, histopathology does not readily distinguish between the two, and the diagnosis of malignancy is made on primarily clinical grounds. Thus, most clinicians feel that the risk of severe hemorrhage outweighs the potential benefits of pretreatment biopsy. However, others argue that the risk of uncontrollable bleeding has been overstated and that biopsy may provide valuable planning information, particularly in cases where the diagnosis is in question and other entities on the differential diagnosis may benefit from a change of treatment plan (51,52). In those cases, biopsy may be most prudently performed as a separate operative procedure, with definitive treatment deferred until permanent histopathology becomes available, or biopsy with intraoperative pathological consultation obtained at the time of resection. Since the possibility of considerable intraoperative hemorrhage can be anticipated for all these lesions, particular attention must be given to preoperative hematologic workup including hemoglobin/hematocrit, coagulation labs, and bleeding time. Strict instructions should be given to avoid non-steroidal anti-inflammatory drugs (NSAIDS) and other blood-thinners for 2 weeks prior to resection, including vitamin E, and herbal preparations containing ginko, ginseng, or
Chapter 34: Angiofibromas and Vascular Tumors of the Skull Base
A
B
C
D
487
Figure 5 Advanced JNA with intracranial extension. Contrast-enhanced CT scan (A–C) and axial T1-weighted MRI with contrast (D). (A) Coronal view. Note invasion of sphenoid sinus, and intracranial penetration (arrow). Although tumor has entered the cranial cavity, it remains extradural (B–C). Axial CT scan viewed in soft tissue (B) and bone (C) windows shows massive invasion of the pterygopalatine fossa with lateral extension to the infratemporal fossa, extensive erosion of the skull base, and pronounced widening of the sphenopalatine foramen (arrows). (D) MRI at approximately the same level highlights the extent of lateral invasion, and pushing, rather than infiltrative, borders. This tumor would be staged as a Radkowski IIIB, and would require a combined head and neck/neurosurgical procedure, including facial translocation, infratemporal fossa, and subtemporal craniectomy approaches. Source: Images courtesy of Laurence Ginsberg, UT MDACC Cancer Center.
garlic. Type-and-cross should be obtained before surgery, and the availability of adequate units of blood should be verified; hypotensive anesthesia is commonly used to minimize blood loss. Autologous blood banking or cell-saver autotransfusion can be considered for large lesions or patients who are likely to poorly tolerate blood loss.
TREATMENT Choice of Treatment Selection of treatment and pretreatment plannings is best accomplished by the surgeon in consultation with a multidisciplinary team experienced in the management of these lesions. In the vast majority of cases, surgery is the preferred management of JNA and other vascular lesions of the skull base. While several series have suggested that primary radiation therapy can achieve control rates comparable to surgery
for JNA, involution of the lesion occurs over the course of months or even years. Radiation therapy also raises the concerns of secondary malignancy (53,54) and alterations of craniofacial growth (55) in this relatively young population. Although extensive intracranial involvement has been argued as an indication for radiation therapy, JNA generally respects the dura and can usually be separated from the intracranial contents without great difficulty. We recommend considering radiation as a primary modality only for patients with tumors that are deemed unresectable or resectable only with great morbidity due to involvement of the cavernous sinus or pituitary, internal carotid, optic chiasm, or optic nerve; or patients who are medically unfit for surgery. Other indications include treatment of persistent or recurrent disease in selected patients. Other therapies, including chemotherapy, hormonal therapy, and definitive embolization, are for the most part unsuitable as frontline therapy, and should be considered
488
Sikora and Weber
A
B
Figure 6 Angiogram of angiofibroma pre- and postembolization. (A) Pre-embolization: the angiofibroma’s primary blood supply is from the internal maxillary artery. (B) Postembolization: there is considerable reduction of tumor vascularity.
experimental. Observation is generally reserved for those unlikely to withstand treatment, since these lesions are unlikely to regress, or (in the case of JNA) unlikely to regress before causing complications.
Surgery Surgical Approaches The goal of surgery is complete removal of the lesion without injury to brain, optic nerve, and other important structures and with minimal hemorrhage. The tendency for these lesions to infiltrate complex and relatively inaccessible anatomical spaces traversed by important neurovascular structures means that failure to meticulously review imaging and to develop an individualized surgical plan virtually guarantees residual/recurrent disease and increased risk for complications. Therefore, there is no single best approach to these lesions; rather, the surgical approach or combination of approaches should be tailored to the anatomical location of the lesion, and surgeons must be comfortable with a wide range of endoscopic and open approaches. An outline for choosing the route of surgical access based on location of the lesion is presented in Table 4. Further information about each approach as it relates to vascular lesions of the skull base is given below; technical details of theses techniques are provided elsewhere in this volume.
Endoscopic Approach Although the appropriateness of endoscopic surgery for JNA has been considered controversial in the past, numerous case series (56–58) as well as growing endoscopic experience with other benign and malignant tumors have suggested that for selected lesions this is a valid approach. The controversy now lies in the extent of disease amenable to endoscopic surgery, with some advocating its use only in lesions limited to the nasal cavity, nasopharynx, and paranasal sinuses, and oth-
ers suggesting that it is appropriate for various degrees of invasion of the pterygopalatine and infratemporal fossae, or even limited intracranial disease (59,60). Advocates of a more aggressive approach to endoscopic resection of JNA cite the ability to closely inspect the resected bed, the generally noninfiltrative nature of the lesion, opportunity to ligate the ipsilateral internal maxillary and sphenopalatine arteries early in the procedure, avoidance of the hemorrhage incurred by approaches requiring osteotomy and extensive soft tissue dissection, and decreased total blood loss as justifications for this approach. Endoscopic approaches also have the obvious advantages of avoiding a facial scar, and avoiding disruption of facial bones and soft tissues, particularly in adolescents who are still undergoing facial growth. We feel that exclusively endoscopic approaches are appropriate for lesions involving the nasal cavity and nasopharynx with limited extension into the pterygomaxillary fissure, so long as the surgeon is already experienced in the techniques of endoscopic sinonasal surgery. Endoscopic approaches may also be used in conjunction with more laterally based open approaches when a lesion requires both anterior and lateral approaches for complete resection.
Transpalatal, Maxillectomy, and Facial Translocation Approaches The transpalatal approach provides potentially good access to medial structures such as the nasopharynx, nasal cavity, and sphenoid; however, it is difficult to get lateral exposure. This procedure also carries the risk of palatal fistula. Although the transpalatal approach was once commonly used for limited JNA, endoscopic techniques provide excellent access to the same areas and in many institutions have replaced the transpalatal approach. Medial maxillectomy, performed with a lateral rhinotomy and with Weber-Ferguson or midfacial degloving
Chapter 34: Angiofibromas and Vascular Tumors of the Skull Base
489
Table 4 Choice of Surgical Access Based on Location of the Lesion Nasal Cavity Transpalatal Nasopharynx | Sphenoid Ethmoid Limited Pterygopalatine Foramen Extensive Pterygopalatine Foramen Limited Infratemporal Fossa Medial Cavernous Sinus Extensive Infratemporal Fossa Lateral Cavernous Sinus Middle Cranial Fossa Anterior Cranial Fossa
| Endoscopic | |
| Infratemporal Fossa Approach | Frontal Craniotomy
incision, provides better access to lateral structures with the potential to reach as far as the infratemporal fossa (ITF). Facial translocation approaches, which allow lateral extension of the exposure afforded by maxillectomy, allow improved access to the ITF and the lateral cavernous sinus (Fig. 7). In general, maxillectomy/transfacial approaches will provide adequate access to the majority of tumors without intracranial involvement. When tumor extends intracranially, the majority of tumors can be resected by combining a transfacial approach with craniotomy (61). Although there are numerous potential complications (such as hemorrhage, injury to the nasolacrimal duct, cheek numbness, and injury to the orbital contents), these procedures are generally well tolerated and significant complications are infrequent. Transfacial approaches share the theoretical concern that disruption of facial development may lead to facial asymmetry and retardation of facial growth in children and adolescents. However, several recent series of craniofacial procedures in children did not find evidence that this was the case (62,63), observations consistent with the experience of Randal S. Weber, the senior author of this article. Thus, while disruption of the facial skeleton should be avoided in young patients where possible, adequate exposure should not be compromised.
Infratemporal Fossa Approach The ITF approaches have been proposed to be versatile procedures for the treatment of extensive JNA, since they allow wide exposure of disease in the lateral reaches of the PPF and ITF, wide exposure of the ICA and cavernous sinus, and exposure of the middle cranial fossa, as well as adequate access to the nasal cavity, nasopharynx, and paranasal sinuses. However, tumor medial to the abducens nerve (VI) is not accessible by this technique, and is usually addressed by postoperative radiation. The type D approach is used for tumors which are not in proximity to the internal carotid artery, whereas the type C approach is used for tumors with medial extension into the ICA or cavernous sinus [reviewed in (64)]. The ITF approaches are extremely morbid, and have the potential for severe complications including hemorrhage, death, CSF leak, brain injury, stroke, malocclusion and Temporomandibular joint (TMJ) dysfunction (due to violation of the TMJ capsule during the approach), and facial nerve (VII) injury (64,65). The type C approach also requires subtotal petrosectomy with ablation of the middle ear cleft and Eustachian tube, leading to severe conductive hearing loss on the operated side (66). Even for advanced JNA, a transfacial approach tailored to the tumor’s location, supplemented with frontal craniotomy when tumors involve the anterior cranial fossa, can address nearly all resectable JNA with less morbidity than the ITF approaches.
| | | Maxillectomy/LeFort | | |
| | | | Transfacial | | |
Complications The potential for complications is proportional to the size, anatomical location of the lesion, and the extent of surgery. For JNA, the most common major complication (besides recurrence) is intraoperative hemorrhage, which can be severe enough to require transfusion, or be life-threatening (67). While steps may be taken to minimize this risk, including preoperative embolization of feeding vessels and intraoperative ligation of the IMA, resection of JNA is an inherently bloody procedure, and the possibility of massive hemorrhage must be planned for. Delayed bleeding is fortunately rare. Intracranial JNA carries the obvious risks of injury to the dura, CSF leak, injury to brain parenchyma, stroke, and other neurological sequelae. Resection of lesions involving, or in close proximity to, the orbit, optic chiasm, or cavernous sinus, carry risks of visual impairment and diplopia. Both infratemporal and transfacial approaches can cause facial numbness due to traction on, or sacrifice of, branches of V. This often improves with time. Virtually all patients should expect some nasal dryness and crusting in the postoperative period. Other complications include sinusitis, serous otitis media, and unfavorable facial scarring. The recurrence rate of JNA has traditionally been much higher than typical for benign disease, and has been estimated in some series to range between 20% and over 50%, with more advanced lesions carrying a significantly higher risk for recurrence (20). Risk factors for recurrence include intracranial disease (especially involvement of the anterior cranial fossa), extension into the basisphenoid through the pterygoid canal, erosion into the walls of the sphenoid sinus, involvement of the medial cavernous sinus, and infiltration into the pterygoid muscles and into the pterygoid plates (20,36). To avoid recurrence, the surgical approach should be carefully tailored to the extent of disease, and meticulous attention to removal of tumor extending into the basisphenoid is recommended.
Radiation Therapy Selected series have found that radiation therapy provides local control rates similar to that of surgery in patients with advanced JNA, and it was formerly considered the treatment of choice for patients with “unresectable” intracranial or cavernous sinus extension. As surgical approaches to JNA have improved, allowing for more aggressive resection of advanced disease with greatly decreased mortality and morbidity, surgery has become the standard of care even of tumors with extensive intracranial spread. However, radiation therapy may still play a role in tumors for which resection would pose unacceptable risk to the optic nerve, optic chiasm, internal carotid artery, and cavernous sinus, particularly when disease is found medial to the abducens (IV) nerve.
490
Sikora and Weber
A
B
C
D
Figure 7 Combined infratemporal fossa and facial translocation approach for advanced angiofibroma with intracranial extension. (A) Hemicoronal and WeberFergusson incisions. (B) Bone cuts for facial translocation, zygomatic osteotomy, and subtemporal craniectomy. (C) Vascularized anterior maxillary osteoplastic flap, reflected. (D) Extent of bone removed in subtemporal craniectomy. Source: Adapted from Ref. 61.
Radiation therapy is also considered for recurrences not amenable to repeat surgery, and after surgery when subtotal resection is necessary to avoid risk to vital structures. It is important to note that radiation therapy usually provides relatively rapid relief of symptoms, but complete involution of the tumor can take years and recurrence is more likely when residual tumor persists more than 2 years (68). Thus, meticulous follow-up with serial imaging is particularly important when patients are treated with this modality. The largest series of JNA patients treated with radiation therapy (55 patients treated with 30–35 Gy) described an 80% control rate (54). More recent series have described similar control rates, including one of 22 patients treated with
definitive radiotherapy (30–36 Gy, conventional fractionation) from 1975 to 2003 which found a 90% control rate at 10 years, with all tumors ultimately controlled by salvage surgery or re-irradiation (69). All patients tolerated therapy without interruptions in treatment; however, there was a significant rate of late complications (32%), including six patients with cataracts, two with in-field basal cell carcinomas, and two with transient central nervous system syndrome. Another series that reported 27 patients treated with definitive radiation therapy for advanced JNA (30–55 Gy) reported a recurrence rate of 15%, with all recurrences occurring within the first 2 to 5 years (70). An additional 15% of patients had severe complications including temporal lobe radionecrosis,
Chapter 34: Angiofibromas and Vascular Tumors of the Skull Base
cataracts, panhypopituitarism, and growth retardation. None of these series reported unfavorable alterations of facial growth, although that is a theoretical concern of radiation in this population. Due to the development of skull base surgery techniques, which allow the reliable and safe resection of advanced JNA, and the risk of significant delayed complications of radiotherapy, we consider surgery frontline therapy for the majority of patients with JNA and reserve radiation therapy for recurrent and residual disease and in combination with therapy to address infiltration of tumor into areas where resection is not desirable. While most published studies are of JNA, treated with conventional radiotherapy, small case series treated with conformal or intensity-modulated radiotherapy have shown similar control rates and an initially favorable complication rate (71,72). However, both patient numbers and duration of follow-up are still too limited to draw conclusions about these relatively new techniques. Several one- and two-patient series of patients with recurrent or residual disease after surgery treated with single-dose stereotactic radiotherapy (“gamma knife” or “cyber knife”) described 100% control (defined as failure of disease to progress) at 2 to 3 years with no complications, but the small number of patients and limited follow-up preclude drawing conclusions from these interesting pilot studies (73,74). For other vascular skull base tumors (HPC, hemangioma, and vascular malformations), radiotherapy has an even less-defined role. Surgery is the usual treatment of these predominantly benign lesions.
Chemotherapy and other Therapies Chemotherapy (with doxorubicin and dacarbazine or Adriamycin and dacarbazine) has been described in one 5-patient series (75), and a single patient who was part of a larger series of chemotherapy for pediatric tumors (76), as a potential treatment for patients with advanced, unresectable disease due to significant intracranial involvement, extensive blood supply from intracranial vessels, or recurrence after adequate surgical treatment. However, it is infrequently used. The presence of hormone receptors on JNA and the obvious hormonal component of its regulation have led to substantial interest in hormonal treatment either alone or to facilitate surgical resection. Estrogen has been shown to decrease the size and vascularity of JNA, but it is impossible to predict which tumors will respond, and the feminizing effects of estrogen are highly undesirable in this patient population. The androgen receptor blocker flutamide has been examined in two small series of five (77)and seven (78) patients; in the first series four patients had an average 44% reduction of tumor volume, and in the second, no reduction in tumor volume or intraoperative blood loss was observed. Currently, the use of chemotherapy and hormonal therapies in the treatment of JNA should be restricted to investigational clinical trials. These therapies have no role in the treatment of other benign vascular skull base lesions.
OUTCOME AND PROGNOSIS The overall prognosis for limited JNA is excellent, with control rates for lesions without intracranial involvement approaching 90 to 100% after surgery alone. Recurrence rates for stage III and IV disease are considerably worse, and historically have ranged from 30% to almost 60% (see discussion of recurrence in section on complications, above) (36,79,80). More recent series suggest that recurrence is lower with mod-
491
ern advanced craniofacial techniques. A report describing a single-center’s experience treating 17 patients with Fisch stage III and IV JNA with preoperative embolization followed by an infratemporal fossa approach, and comparing results in this group to 16 additional patients previously reported by the same institution, found a recurrence rate of 6% (1 patient) after a median 28-month follow-up, despite the fact that 8 patients had had previous operations (64). Complications included wound infection (6%), facial palsy involving the frontal branch (25%, with 4/5 recovering by 6 months), and permanent conductive hearing loss in all patients treated with a type C approach; one patient had partial vision loss after embolization. There was no incidence of death, stroke, CSF leak, meningitis, or injury to the brain parenchyma. The senior author’s experience using preoperative embolization followed by a craniofacial approach to treat five patients with advanced (Radkowski stage IIIB) JNA found one recurrence (20%) after 28 to 63 months of follow-up (61). Complications were minimal, including nasal crusting and discharge, sinusitis, temporary serous otitis media (60%), and facial anesthesia in the distribution of V2 and V3. Less than 50% of recurrences of JNA occur within 1 year of surgery (20), and recurrences can present at any time before young adulthood. Thus, cautious posttreatment follow-up, including serial endoscopic and CT and/or MRI imaging, is necessary to detect and treat recurrence before it results in complications. We recommend physical examination and imaging every 3 months during the first year, every 6 months during the second year following surgery, and yearly examinations for the next 3 years. Recurrence is extremely unlikely after 5 years of uneventful follow-up. The prognosis of other vascular skull base lesions is hard to estimate due to their rarity. For HPC, the local control recurrence rate has been estimated to be 40%, with metastases developing in an additional 15% of patients; 10-year survival is estimated to be 70% for HPC at all sites (81). As is the case with JNA, sinonasal HPC demands close endoscopic and radiographic follow-up after treatment. Clinicians should be mindful of the potential for metastases to lung, liver, and bone even in disease with apparently benign behavior, and perform symptom-directed evaluation of these areas as necessary.
REFERENCES 1. Batsakis JG. Tumors of the Head and Neck: Clinical and Pathological Correlations, 2nd ed. Baltimore, MD: Williams and Wilkins, 1979. 2. Bremer JW, et al. Angiofibroma: Treatment trends in 150 patients during 40 years. Laryngoscope. 1986;96(12):1321–1329. 3. Gatumbi I, Linsell CA. Nasopharyngeal angiofibromas in kenya. Br J Cancer. 1964;18:69–73. 4. Backwinkel KD, Diddams JA. Hemangiopericytoma. Report of a case and comprehensive review of the literature. Cancer. 1970;25:896–901. 5. Walike JW, Bailey BJ. Head and neck hemangiopericytoma. Arch Otolaryngol. 1971;93:345–353. 6. Mulliken JB, Glowacki J. Hemangiomas and vascular malformations in infants and children: A classification based on endothelial characteristics. Plast Reconstr Surg. 1982;69(3):412–422. 7. Jacobsson M, et al. Involution of juvenile nasopharyngeal angiofibroma with intracranial extension. A case report with computed tomographic assessment. Arch Otolaryngol Head Neck Surg. 1989;115(2):238–239. 8. Stansbie JM, Phelps PD. Involution of residual juvenile nasopharyngeal angiofibroma (a case report). J Laryngol Otol. 1986;100(5):599–603.
492
Sikora and Weber
9. Abraham SC, et al. Frequent beta-catenin mutations in juvenile nasopharyngeal angiofibromas. Am J Pathol. 2001;158(3):1073– 1078. 10. Valanzano R, et al. Genetic evidence that juvenile nasopharyngeal angiofibroma is an integral FAP tumour. Gut. 2005;54(7):1046–1047. 11. Guertl B, et al. Nasopharyngeal angiofibroma: An APC-geneassociated tumor? Hum Pathol. 2000;31(11):1411–1413. 12. Ferouz AS, Mohr RM, Paul P. Juvenile nasopharyngeal angiofibroma and familial adenomatous polyposis: An association? Otolaryngol Head Neck Surg. 1995;113(4):435–439. 13. Giardiello FM, et al. Nasopharyngeal angiofibroma in patients with familial adenomatous polyposis. Gastroenterology. 1993;105(5):1550–1552. 14. Farag MM, et al. Hormonal receptors in juvenile nasopharyngeal angiofibroma. Laryngoscope. 1987;97(2):208–211. 15. Hwang HC, et al. Expression of androgen receptors in nasopharyngeal angiofibroma: An immunohistochemical study of 24 cases. Mod Pathol. 1998;11(11):1122–1126. 16. Brentani MM, et al. Multiple steroid receptors in nasopharyngeal angiofibromas. Laryngoscope. 1989;99(4):398–401. 17. Gatalica Z. Immunohistochemical analysis of steroid hormone receptors in nasopharyngeal angiofibromas. Cancer Lett. 1998;127(1–2):89–93. 18. Neel HB III, et al. Juvenile angiofibroma. Review of 120 cases. Am J Surg. 1973;126(4):547–556. 19. Harrison DF. The natural history, pathogenesis, and treatment of juvenile angiofibroma. Personal experience with 44 patients. Arch Otolaryngol Head Neck Surg. 1987;113(9):936–942. 20. Lloyd G, et al. Juvenile angiofibroma: The lessons of 20 years of modern imaging. J Laryngol Otol. 1999;113(2):127–134. 21. Wilms G, et al. Pre-operative embolization of juvenile nasopharyngeal angiofibromas. J Belge Radiol. 1989;72(6):465–470. 22. Sternberg SS. Pathology of juvenile nasopharyngeal angiofibroma; a lesion of adolescent males. Cancer. 1954;7(1):15–28. 23. Svoboda DJ, Kirchner F. Ultrastructure of nasopharyngeal angiofibromas. Cancer. 1966;19(12):1949–1962. 24. Hasegawa T, et al. Solitary fibrous tumor of the soft tissue. An immunohistochemical and ultrastructural study. Am J Clin Pathol. 1996;106(3):325–331. 25. Alobid I, et al. Solitary fibrous tumour of the nasal cavity and paranasal sinuses. Acta Otolaryngol. 2003;123(1):71–74. 26. Thompson LD, Miettinen M, Wenig BM. Sinonasal-type hemangiopericytoma: A clinicopathologic and immunophenotypic analysis of 104 cases showing perivascular myoid differentiation. Am J Surg Pathol. 2003;27(6):737–749. 27. Stout AP, Murry MR. Hemangiopericytoma: A vascular tumor featuring Zimmermann’s pericytes. Ann Surg. 1942;116:26–33. 28. Gengler C, Guillou L. Solitary fibrous tumour and haemangiopericytoma: Evolution of a concept. Histopathology. 2006;48(1):63–74. 29. Compagno J. Hemangiopericytoma-like tumors of the nasal cavity: A comparison with hemangiopericytoma of soft tissues. Laryngoscope. 1978;88(3):460–469. 30. Carew JF, Singh B, Kraus DH. Hemangiopericytoma of the head and neck. Laryngoscope. 1999;109(9):1409–1411. 31. Batsakis JG, Rice DH. The pathology of head and neck tumors: Vasoformative tumors, part 9B. Head Neck Surg. 1981;3(4):326– 339. 32. Chang J, et al. Proliferative hemangiomas: Analysis of cytokine gene expression and angiogenesis. Plast Reconstr Surg. 1999;103(1):1–9; discussion 10. 33. Elluru RG, Azizkhan RG. Cervicofacial vascular anomalies. II. Vascular malformations. Semin Pediatr Surg. 2006;15(2):133– 139. 34. Maher ER, Kaelin WG Jr. von Hippel-Lindau disease. Medicine (Baltimore). 1997;76(6):381–391. 35. Werner JA, et al. Current concepts in the classification, diagnosis and treatment of hemangiomas and vascular malformations of the head and neck. Eur Arch Otorhinolaryngol. 2001;258(3):141– 149. 36. Radkowski D, et al. Angiofibroma. Changes in staging and treatment. Arch Otolaryngol Head Neck Surg. 1996;122(2):122–129.
37. Sessions RB, et al. Radiographic staging of juvenile angiofibroma. Head Neck Surg. 1981;3(4):279–283. 38. Yeh S, Foroozan R. Orbital apex syndrome. Curr Opin Ophthalmol. 2004;15(6):490–498. 39. Schick B, Kahle G. Radiological findings in angiofibroma. Acta Radiol. 2000;41(6):585–593. 40. Chong VF, Fan YF. Radiology of the nasopharynx: Pictorial essay. Australas Radiol. 2000;44(1):5–13. 41. Vilanova JC, Barcelo J, Villalon M. MR and MR angiography characterization of soft tissue vascular malformations. Curr Probl Diagn Radiol. 2004;33(4):161–170. 42. Li JR, et al. Evaluation of the effectiveness of preoperative embolization in surgery for nasopharyngeal angiofibroma. Eur Arch Otorhinolaryngol. 1998;255(8):430–432. 43. Siniluoto TM, et al. Value of pre-operative embolization in surgery for nasopharyngeal angiofibroma. J Laryngol Otol. 1993;107(6):514–521. 44. Moulin G, et al. Juvenile nasopharyngeal angiofibroma: Comparison of blood loss during removal in embolized group versus nonembolized group. Cardiovasc Intervent Radiol. 1995;18(3):158–161. 45. George B, et al. Intratumoral embolization of intracranial and extracranial tumors: Technical note. Neurosurgery. 1994;35(4):771– 773; discussion 773–774. 46. Liang Y, et al. Direct intratumoral embolization of hypervascular tumors of the head and neck. Chin Med J (Engl). 2003;116(4):616–619. 47. da Costa DM, et al. Surgical experience with juvenile nasopharyngeal angiofibroma. Ann Otolaryngol Chir Cervicofac. 1992;109(5):231–234. 48. Duvall AJ III, Moreano AE. Juvenile nasopharyngeal angiofibroma: Diagnosis and treatment. Otolaryngol Head Neck Surg. 1987;97(6):534–540. 49. Jackson IT, et al. Hemangiomas, vascular malformations, and lymphovenous malformations: Classification and methods of treatment. Plast Reconstr Surg. 1993;91(7):1216–1230. 50. Hovius SE, et al. The diagnostic value of magnetic resonance imaging in combination with angiography in patients with vascular malformations: A prospective study. Ann Plast Surg. 1996;37(3):278–285. 51. Scholtz AW, et al. Juvenile nasopharyngeal angiofibroma: management and therapy. Laryngoscope. 2001;111(4 Pt 1):681–687. 52. Burkey B, Koopmann CF, Brunberg J. The use of biopsy in the evaluation of pediatric nasopharyngeal masses. Int J Pediatr Otorhinolaryngol. 1990;20(2):169–179. 53. Cummings BJ. Relative risk factors in the treatment of juvenile nasopharyngeal angiofibroma. Head Neck Surg. 1980;3(1):21– 26. 54. Cummings BJ, et al. Primary radiation therapy for juvenile nasopharyngeal angiofibroma. Laryngoscope. 1984;94(12 Pt 1):1599–1605. 55. Denys D, et al. The effects of radiation on craniofacial skeletal growth: A quantitative study. Int J Pediatr Otorhinolaryngol. 1998;45(1):7–13. 56. Roger G, et al. Exclusively endoscopic removal of juvenile nasopharyngeal angiofibroma: Trends and limits. Arch Otolaryngol Head Neck Surg. 2002;128(8):928–935. 57. Hofmann T, et al. Endoscopic resection of juvenile angiofibromas–long term results. Rhinology. 2005;43(4):282– 289. 58. Pryor SG, Moore EJ, Kasperbauer JL. Endoscopic versus traditional approaches for excision of juvenile nasopharyngeal angiofibroma. Laryngoscope. 2005;115(7):1201–1207. 59. Onerci TM, Yucel OT, Ogretmenoglu O. Endoscopic surgery in treatment of juvenile nasopharyngeal angiofibroma. Int J Pediatr Otorhinolaryngol. 2003;67(11):1219–1225. 60. Sciarretta V, et al. Endoscopic sinus surgery for the treatment of vascular tumors. Am J Rhinol. 2006;20(4):426–431. 61. Bales C, et al. Craniofacial resection of advanced juvenile nasopharyngeal angiofibroma. Arch Otolaryngol Head Neck Surg. 2002;128(9):1071–1078. 62. Powell DM, et al. Maxillary removal and reinsertion in pediatric patients. Arch Otolaryngol Head Neck Surg. 2002;128(1):29–34.
Chapter 34: Angiofibromas and Vascular Tumors of the Skull Base 63. Lang DA, et al. Craniofacial access in children. Acta Neurochir (Wien). 1998;140(1):33–40. 64. Zhang M, et al. Update on the infratemporal fossa approaches to nasopharyngeal angiofibroma. Laryngoscope. 1998;108(11 Pt 1):1717–1723. 65. Fisch U, Fagan P, Valavanis A. The infratemporal fossa approach for the lateral skull base. Otolaryngol Clin North Am. 1984;17(3):513–552. 66. Fisch U, Pillsbury HC. Infratemporal fossa approach to lesions in the temporal bone and base of the skull. Arch Otolaryngol. 1979;105(2):99–107. 67. Tyagi I, Syal R, Goyal A. Staging and surgical approaches in large juvenile angiofibroma–study of 95 cases. Int J Pediatr Otorhinolaryngol. 2006;70(9):1619–1627. 68. Reddy KA, et al. Long-term results of radiation therapy for juvenile nasopharyngeal angiofibroma. Am J Otolaryngol. 2001;22(3):172–175. 69. McAfee WJ, et al. Definitive radiotherapy for juvenile nasopharyngeal angiofibroma. Am J Clin Oncol. 2006;29(2):168–170. 70. Lee JT, et al. The role of radiation in the treatment of advanced juvenile angiofibroma. Laryngoscope. 2002;112(7 Pt 1):1213–1220. 71. Kuppersmith RB, et al. The use of intensity modulated radiotherapy for the treatment of extensive and recurrent juvenile angiofibroma. Int J Pediatr Otorhinolaryngol. 2000;52(3):261– 268. 72. Beriwal S, Eidelman A, Micaily B. Three-dimensional conformal radiotherapy for treatment of extensive juvenile angiofibroma: report on two cases. ORL J Otorhinolaryngol Relat Spec. 2003;65(4):238–241.
493
73. Park CK, et al. Recurrent juvenile nasopharyngeal angiofibroma treated with gamma knife surgery. J Korean Med Sci. 2006;21(4):773–777. 74. Dare AO, et al. Resection followed by radiosurgery for advanced juvenile nasopharyngeal angiofibroma: Report of two cases. Neurosurgery. 2003;52(5):1207–1211; discussion 1211. 75. Goepfert H, Cangir A, Lee YY. Chemotherapy for aggressive juvenile nasopharyngeal angiofibroma. Arch Otolaryngol. 1985;111(5):285–289. 76. Cangir A, et al. Combination chemotherapy with adramycin (NSC-123127) and dimethyl triazeno imidazole carboxamide (DTIC) (NSC-45388) in children with metastatic sold tumors. Med Pediatr Oncol. 1976;2(2):183–190. 77. Gates GA, et al. Flutamide-induced regression of angiofibroma. Laryngoscope. 1992;102(6):641–644. 78. Labra A, et al. Flutamide as a preoperative treatment in juvenile angiofibroma (JA) with intracranial invasion: Report of 7 cases. Otolaryngol Head Neck Surg. 2004;130(4):466–469. 79. Gullane PJ, et al. Juvenile angiofibroma: A review of the literature and a case series report. Laryngoscope. 1992;102(8):928– 933. 80. Howard DJ, Lloyd G, Lund V. Recurrence and its avoidance in juvenile angiofibroma. Laryngoscope. 2001;111(9):1509–1511. 81. Hasson O, Kirsch G, Lustmann J. Hemangiopericytoma of the tongue in an 11-year-old girl: Case report and literature review. Pediatr Dent. 1994;16(1):49–52. 82. Fisch U. The infratemporal fossa approach for nasopharyngeal tumors. Laryngoscope. 1983;93(1):36–44.
35 Chordoma and Chondrosarcoma of the Skull Base Gordon T. Sakamoto and Griffith R. Harsh
chondrocranium (14). The petroclival synchondrosis is the most common site (15). Chondrosarcomas are also extradural tumors. Most grow slowly, destroy bone, and extend into surrounding soft tissue. Distant metastases occur in 7% to 12% of patients (14,16). Most chondrosarcomas are low-grade malignancies (15) and thus have a better prognosis than chordomas (17,18).
INTRODUCTION Chordomas and chondrosarcomas of the skull base can present formidable challenges to effective treatment. These challenges include the relative inaccessibility of the skull base, tumor involvement of critical neural and vascular structures, a tendency to recur locally, and possible metastatic dissemination. Despite such challenges, the combination of aggressive skull base surgical resection and high-dose radiation can cure almost all low-grade chondrosarcomas and provide meaningful intervals of disease control of both chordomas and high-grade chondrosarcomas.
INCIDENCE AND EPIDEMIOLOGY Chordomas
Chordomas were first recognized at autopsy by Lushka in 1856 (1) and Virchow in 1857 (2). Believing that these tumors were cartilaginous in origin, Virchow named them “ecchondrosis physaliphora” (2). In 1858, Muller proposed that these tumors were related to the notochord (3). In 1864, Klebs described the first symptomatic case (4). In 1894, Ribbert coined the term “chordoma” after he found these tumors in the nucleus pulposus and correctly surmised their notochordal origin (5). Chordomas are rare primary malignant tumors of bone that arise from notochordal remnants (6). They are locally aggressive. Although they can occur anywhere along the axial skeleton, chordomas most commonly occur at either end, the sacrococcyx (50%) or the clivus (35%) (6,7). Most grow slowly, expanding and destroying bone. Although chordomas usually arise outside the dura, they may infiltrate and penetrate the dura to spread intracranially or intraspinally. Dural invasion usually occurs late in the course of aggressive tumors. Chordomas can also extend intradurally through surgical durotomies. There are rare reports of primary intradural intracranial chordomas (8,9). Metastases, which become clinically evident in 10% to 20% of patients, usually occur late in the course of the disease (10). At autopsy, metastases can be found in up to 40% of patients (10). Patients usually die from the consequences of locoregional disease, rather than from the metastases.
Chordomas are the most common extradural clival tumors. The overall incidence of chordomas is less than 0.1 per 100,000 persons per year (19), and they account for about 0.15% of all intracranial tumors (20). Skull base chordomas have an equal gender distribution (21–23). Although chordomas can occur at any age, they are rare in patients younger than 30 years old (24). The peak incidence is in the fourth or fifth decade of life (25). The median age at diagnosis in a large series was 46 years (22). There is no known association between the development of chordomas and potential risk factors such as radiation or other environmental carcinogens. Chordomas occur in isolation and are not part of any known systemic syndrome. Although chordoma occurrence in one family has been linked to chromosome 7q33 (26), no gene mutation specific to chordoma has been identified. Intracranial chordomas most commonly arise from the midline caudal third of the clivus, below the spheno-occipital synchondrosis (22,27). Chordomas may extend in all directions from their notochordal origin (27). Chordomas from the rostral notochord often extend into the dorsum sellae and present as sellar, suprasellar, or cavernous sinus tumors, which compress the pituitary gland, optic nerves and chiasm, and the midbrain (28). Chordomas extending ventrally through the clivus can present as nasopharyngeal masses causing nasal obstruction or dysphasia (29). Chordomas extending from the dorsal clivus can compress the pons and medulla; dorsolaterally extending tumors can involve the spheno-occipital or petrosal temporal bone.
Chondrosarcomas
Chondrosarcomas
Chondrosarcoma was first recognized as a distinct pathologic entity in 1939 by the American College of Surgeons (11). Previously, these tumors were classified as osteosarcomas. Chondrosarcomas are rare cartilaginous tumors of different grades of malignancy. Their origin is controversial; possibilities include embryonal cartilaginous rests, mesenchymal pluripotent cells, and metaplasia of fibroblasts (12,13). Approximately 50% of chondrosarcomas occur at skull base synchondroses, sites of fusion of separate cartilages forming the
Chondrosarcomas are rare. They account for 0.02% of all intracranial neoplasms (7,14,30). Chondrosarcomas are slightly more common in men than in women (16). Although skull base chondrosarcomas can occur in any age group, there is a peak incidence in the second and third decades (14,31). The mean age at diagnosis is 40.7 years (13). Although chondrosarcomas are usually isolated tumors, they may occur as part of a systemic syndrome, such as Paget disease, Ollier disease, and Maffucci syndrome (15,32). Ollier disease involves
Chordomas
495
496
Sakamoto and Harsh
multiple enchondral bone cysts and Maffucci syndrome involves multiple enchondromas and cutaneous and visceral hemangiomas. Ploidy ranges from hyperhaploidy to pentaploidy (33). Although loss (and gain) of genetic material from many different chromosomes occurs, cytogenetic studies have identified no single characteristic aberration (33). Isolated chondrosarcomas are usually paramedian. However, when part of Ollier or Maffucci syndrome, chondrosarcomas may be midline (34). Embryologically, the skull base forms from a cartilaginous matrix (35). During ossification, some cartilage may fail to form bone and remain as a rest. These cartilaginous rests may transform into chondrosarcomas. Although chondrosarcomas can arise from cartilage forming the anterior, middle, or posterior fossa, most skull base chondrosarcomas arise near the clivus. Sixty-six percent arise from the petro-occipital junction, 28% from the clivus and 6% from the sphenoethmoid complex (15).
PATHOLOGY Chordomas Chordomas are lobulated, grayish tan to bluish white tumors. They are usually well delineated from surrounding soft tissues. They range in consistency from firm to gelatinous. There may be foci of calcification or hemorrhage. The tumor’s size varies greatly. In bone, the tumor infiltrates the marrow space and expands the cortex to form a well-demarcated mass. The tumor may also penetrate the cortex and grow into neighboring soft tissue. Chordoma are composed of lobules and nests of large epithelial-appearing cells separated by fibrous bands. The neoplastic cells are arranged in sheets or cords or float individually in the myxoid stroma. The nuclei are of moderate size and show mild to moderate atypia. They have abundant pink cytoplasm. Variable numbers of cells have clear vacuoles, which impart a “bubbly” appearance to the cytoplasm (Fig. 1). Mitoses are limited and foci of necrosis are common. The neoplastic cells contain periodic acid-Schiff diastase-sensitive glycogen. Immunohistochemically, chordomas express S-100 and epithelial markers such as cytokeratin and epithelial membrane antigen (6).
Figure 1 H & E stained chordoma at 200×. Physaliferous cells with multiple clear cytoplasmic vacuoles are arranged in chords.
Chondrosarcomas Macroscopically, chondrosarcomas consist of gray-to-tan white nodules. The tumor consistency ranges from firm and gritty to mucinous. Additionally, there may be yellow-white chalky areas of calcification. They infiltrate the normal marrow and encase cancellous bone. They may transgress the cortex and form a soft tissue mass. Microscopically, four primary types of chondrosarcoma have been described: conventional, clear cell, dedifferentiated, and mesenchymal (15,36). Almost all skull base chondrosarcomas are of the conventional type. The dedifferentiated and mesenchymal variants are more aggressive tumors (25) and rarely affect the skull base. Conventional chondrosarcoma is composed of hyaline, myxoid, or a combination of hyaline and myxoid cartilage. Mixed hyaline and myxoid chondrosarcomas contain variable amounts of both matrices. Hyaline chondrosarcomas (Fig. 2) are characterized by hypercellular hyaline cartilage. The neoplastic chondrocytes lie in clear lacunae within the hyaline matrix. The chondrocytes vary in size and shape. The chondrocyte nuclei have fine chromatin and small nucleoli and vary in size and shape from small and round to medium size and ovoid. The cytoplasm may be clear or eosinophilic. The cytoplasm may also have a bubbly appearance which mimics that of the physaliphorous cells of a chordoma (15). Mitotic activity is usually very low and foci of necrosis may be present. Myxoid chondrosarcomas have neoplastic cells that appear to float in a mucinous matrix. The tumor cells may be bipolar or stellate. The cells are typically arranged in a honeycomb network of interconnecting strands and cords of cells. Mitoses are rare. Immunohistochemically, conventional chondrosarcomas express vimentin and S-100, as do chordomas (15). However, chondrosarcomas typically do not express epithelial markers such as keratin and epithelial membrane antigen (15). This is useful in distinguishing low-grade chondrosarcomas from chordomas.
Grading Grading is important in chondrosarcomas, as it has been shown to predict prognosis (37). Conventional
Figure 2 H & E stained hyaline chondrosarcoma at 100×. There are slightly atypical cells within the lacunar spaces of the hyaline matrix.
Chapter 35: Chordoma and Chondrosarcoma of the Skull Base
chondrosarcomas are graded according to the degree of cellularity, atypia, and mitotic activity (38). Most grading systems employ a three- or four-tiered system. Most chondrosarcomas are well to moderately differentiated. In a series of 200 skull base chondrosarcomas, 50.5% were grade I, 28.5% were a mixture of grade I and grade II, and 21% were pure grade II tumors. There were no grade III tumors (15). Grade I chondrosarcomas are very similar to enchondromas. However, the cellularity is higher, and there is mild cellular pleomorphism. The nuclei are small but often have an open chromatin pattern and small nucleoli. Mitoses are very rare. Grade I chondrosarcomas usually do not metastasize. Grade II chondrosarcomas have higher cellularity than do grade I tumors. They also have moderate cellular pleomorphism, larger nuclei, and more nuclear atypia. Mitoses are rare. Unlike grade I tumors, about 10% to 15% of grade II chondrosarcomas metastasize. Grade III chondrosarcomas have high cellularity, marked cellular pleomorphism, a high nuclear to cytoplasm ratio, and frequent mitoses. Grade III tumors have significant metastatic potential.
CLINICAL PRESENTATION The presentation of chordomas is variable and can be influenced by the tumor’s location and the patient’s age. Symptoms can manifest by compression or invasion of neural tissue. Clival tumors commonly present with pain and cranial neuropathies (6,13,39). The headaches are usually occipital and are worsened by changes in neck position. The cranial neuropathies vary with tumor location: upper clival lesions may affect cranial nerves II–VI, and lower clival lesions may affect cranial nerves VI–XII. Myelopathy may be caused by tumors compressing the brain stem or upper cervical spinal cord. Diplopia and headache are the most common symptoms of skull base chordomas (6,39). In one series, more than 90% of 155 skull base chordomas presented with diplopia (6), in most cases secondary to a sixth nerve palsy. Fifty percent of patients had headache. Fifty percent had one or more palsies of cranial nerves VII through XII. Twenty-five percent had loss of visual acuity and diplopia, and 25% had nasal involvement (6). Additionally, pituitary dysfunction, seizures, and fifth nerve palsy occur (6,13,39). Diplopia (64%), tongue weakness (60%), and headache (45%) were the most common signs and symptoms in one series of pediatric skull base chordomas (40). In children 5 years or younger, long tract signs (88%), lower cranial nerve deficits (62%), and signs of increased intracranial pressure (50%) were the most frequent presentations (40). The presentation of chondrosarcomas is very similar to that of chordomas. Patients commonly present with cranial nerve palsies, headaches, and gait disturbance (15). In a mixed series of skull base chondrosarcomas and chordomas, diplopia (60%), headache (60%), dysphagia (40%), and facial numbness (33%) were the most common symptoms (13). Tumor compression of the brain stem and fourth ventricle can cause lower cranial neuropathy (31), gait ataxia, and increased intracranial pressure (14).
DIFFERENTIAL DIAGNOSIS Most tumors of the skull base arise from its bone or surrounding soft tissue (41,42). Skull base tumors other than chordoma or chondrosarcoma include chondroma, craniopharyngioma, eosinophilic granuloma, fibrous dysplasia, giant cell
497
tumor, lymphoma, meningioma, metastases, osteoblastoma, pituitary adenoma, and plasmacytoma/multiple myeloma. Chondromas are benign tumors composed of mature hyaline cartilage. They grow and usually become symptomatic during adolescence. Plasmacytomas and multiple myeloma are malignant plasma cell tumors that grow within bone. Pituitary adenomas arise within the sella, which is usually expanded. Lymphomas are more likely to involve adjacent soft tissue. In fibrous dysplasia, normal bone matrix is replaced with abnormal calcified tissue containing collagen and fibroblasts (43). Fibrous dysplasia normally presents in late childhood or adolescence. The monostotic form is more common; the polyostotic form is associated with Albright syndrome. Meningiomas in this region can arise from the dura of the clivus (0.6–0.8%) (44), sella, cavernous sinus, petrous apex, or the foramen magnum (2–3%) (42,44). They occur more frequently in women. Clival meningiomas may present with cranial nerve palsies or myelopathy. Foramen magnum meningiomas may also cause local pain. Eosinophilic granulomas usually occur during childhood and present as an enlarging tender mass that appears lytic without a rim of sclerosis on skull X-rays or computed tomography (CT) scan (45). Eosinophilic granulomas rarely affect the skull base, but those that do can cause otorrhea and cranial nerve palsies (45). Osteomas and osteoblastomas are blastic lesions that rarely involve the skull base. Osteoid osteomas are smaller than osteoblastomas and present with pain that is relieved by aspirin. Osteoblastomas present with pain that is often nocturnal. Metastases to the bone, particularly those from breast, lung, and prostate cancer, are quite common.
IMAGING Imaging is critical to the diagnosis and monitoring of chordomas and chondrosarcomas. Chordomas of the skull base are usually found in the midline arising from the clivus (Fig. 3). On noncontrast CT, they are usually isodense or slightly hypodense to the brain. They expand and erode bone (46). Chordomas show moderate-to-marked contrast enhancement on CT. On magnetic resonance imaging (MRI), chordomas are isointense to brain on T1-weighted images, hyperintense on T2 images, and enhance with gadolinium. Blood products from tumoral hemorrhage may change the tumor’s signal characteristics. Chondrosarcomas of the skull base are usually paramedian (Fig. 4) at the petro-occipital junction (66%), the clivus (28%), and the sphenoethmoid complex (6%) (15). Contrast CT shows calcified tumor destroying bone (47). Solid portions of the tumor enhance with contrast. On MRI, chondrosarcomas are hypointense to brain on T1-weighted images and hyperintense on T2-weighted, proton-density, and FLAIR images (48). Chondrosarcomas show heterogeneous contrast enhancement, which may give them a salt and pepper appearance.
TREATMENT Options for the management of skull base chordomas and chondrosarcomas include clinical and radiological observation, biopsy followed by observation, biopsy followed by radiation, surgical removal, and surgery followed by radiation. In addition to these options, chemotherapy has also been used. Although these tumors have characteristic appearances on CT and MRI, the differential diagnosis (Table 1)
498
Sakamoto and Harsh Table 1 Differential Diagnosis Chondroma Craniopharyngioma Eosinophilic granuloma Fibrous dysplasia Lymphoma Meningioma Metastasis Nasopharyngeal carcinoma Neurofibroma Pituitary adenoma Plasmacytoma/Multiple myeloma
(A)
(B)
(C)
(D)
logic function. The goals of surgical resection may vary: en bloc excision of tumor, piecemeal gross total resection of the tumor, or a radical subtotal removal to decompress critical neurovascular structures or to improve tumor geometry for postoperative radiotherapy. The surgical approach should be tailored to the goal chosen for each individual patient. Choice of surgical approach should also consider tumor size, site of origin, direction of expansion, relationships with cranial nerves and arteries, extent of tumor invasion, the patient’s preoperative health, the surgeon’s familiarity with the approach, and prior treatments. Many of surgical approaches to the central skull base have been described. Schematically, they can be broken down into three general approaches: anterior, anterolateral, and
Figure 3 Chordoma. (A) Contrast-enhanced CT shows a midclival mass destroying the clivus and extending posteriorly to compress the basilar artery. (B) T1-weighted post gadolinium axial MR image shows this chordoma to be moderately enhancing. (C) T2-weighted axial MR image shows the chordoma to be hyperintense to brain and hypointense to CSF. (D) T1-weighted coronal image shows destruction of the clivus by a midline tumor.
of these skull base tumors is usually broad enough to warrant tissue examination. Since these lesions are predominantly extradural, standard stereotactic biopsy techniques may not be useful and biopsy is more likely to be performed through the nose, mouth, or mastoid.
(A)
(B)
(C)
(D)
Surgery When the diagnosis is firmly established as chordoma or chondrosarcoma, surgery will usually play a prominent role in treatment. Surgery can provide a definitive pathologic diagnosis, improve neurologic function by decompressing critical structures, lengthen time to recurrence and patient survival, and optimize spatial relationships and tumor geometry for postoperative radiotherapy. Surgical resection is usually indicated in skull base chordomas and chondrosarcomas. Some low-grade chondrosarcomas may be monitored if they are small and asymptomatic, but most chondrosarcomas and all chordomas warrant treatment. The multidisciplinary treatment team of neurosurgeon, otolaryngologist, and radiation oncologist should develop a comprehensive treatment plan which maximizes tumor control and patient survival while limiting the risk of iatrogenic complications. Numerous reports suggest the value to both tumor control and patient survival of extensive resection and high-dose radiotherapy for both chordomas and chondrosarcomas (13,17,18,22,49). Surgery is also often indicated for the restoration or preservation of neuro-
Figure 4 Chondrosarcoma. (A) On a T1-weighted axial MR image, this chondrosarcoma is isointense to brain. (B) Postgadolinium MR imaging shows the heterogeneous “salt and pepper” enhancement of the tumor. (C) A T2weighted axial MR image shows the tumor to be hyperintense to brain. (D) A contrast-enhanced CT demonstrates calcification and enhancement of the tumor.
Chapter 35: Chordoma and Chondrosarcoma of the Skull Base Table 2 Operative Approaches to Chordomas and Chondrosarcomas Anterior approaches Extended subfrontal (modified transbasal) Maxillotomy and extended maxillotomy Transethmoidal Transsphenoidal and extended transsphenoidal Transoral Anterolateral approaches Frontotemporal Lateral Approaches Frontotemporal transsylvian Subtemporal anterior transpetrosal Subtemporal infratemporal preauricular
499
defect should be obliterated by an inlay graft whose edges are placed deep to the margins of the dura. Repaired dura should then be covered by an onlay graft of fascia or dural allograft held snugly by a firm strut (thin bone, polyethylene glycol plate, or titanium wire mesh) wedged beneath the deep margins of the bone opening. This can then be covered by tissue adhesive, fat, and vascularized soft tissue such as nasal mucosa, pharyngeal musculature, pericranium, or temporalis fascia or muscle. With large, insecurely repaired openings, drainage of CSF through a ventriculostomy or lumbar drain is indicated. At the craniocervical junction, the combination of tumor erosion and surgical removal of involved bone may cause instability, which warrants fusion. Instrumentation is then often needed because most patients will receive radiotherapy, which will retard bone fusion.
Posterolateral approaches Extreme lateral Presigmoid petrosal Retrolabyrinthine or partial translabyrinthine
Radiation
posterolateral Table 2. A combination of approaches may be required to obtain satisfactory tumor exposure, especially for larger tumors. Microscopic and endoscopic visualization may both be helpful (50). Since most chordomas and chondrosarcomas are predominantly extradural, an extradural approach is usually preferred (51). This allows direct access to affected bone. Often involved bone is removed in the approach to infiltrated dura. Examples include a transclival approach to a chordoma compressing the pons and a transpetrosal approach to a petroclival chondrosarcoma. Chondrosarcomas tend to be more discrete and thus more completely resectable than chordomas. Some chondrosarcomas, however, may be so heavily calcified and incorporated into the skull base that they can be removed only by fragmenting them into dense parcels or drilling. In manipulating these calcified fragments, it is essential to first identify and then dissect tumor from nearby cranial nerves and critical vessels. Incorporation of fine cranial nerves or vessels within a calcified mass may preclude safe tumor removal and warrant subtotal resection. This portion of the tumor is often relatively indolent and long-term tumor control rates with adjuvant radiotherapy are quite high. Chordomas, grossly, often have two intermixed components: a soft, gelatinous portion within expanded bone or dura and a more sinewy infiltration and expansion of dura or extracranial soft tissue. The gelatinous part can be easily removed by suction or gentle dissection and curettage; often thickened arachnoid protects cranial nerves, brain stem, and vessels. The exception occurs in reoperations in which this protection has been violated by prior dissection. Then, tumor may surround cranial nerves and extend between brain stem arteries and the pia. Inappropriately aggressive resection then risks cranial neuropathies and brain stem stroke from injury to perforating arteries. The sinewy portion requires more sharp dissection. The plane between tumor and surrounding soft tissue is often obscure. Where possible, thickened, potentially infiltrated dura and extradural soft tissue should be excised. After dural and intradural tumor has been removed, any remaining tumor-infiltrated bone at the margins of resection should be aggressively drilled away. The result is usually a large fistula through dura and bone, which must be repaired to prevent leakage of CSF. Dural openings should be sutured closed, primarily if possible, or by using autologous fascia or dural allograft if needed. If a graft cannot be sewn in, the
Few chondrosarcomas and almost no chordomas will be removed with such confidence of microscopically complete resection that adjuvant therapy can be eschewed. Incomplete resection is usually followed by regrowth of tumor (6,13,23,52,53). High-dose radiation can delay or prevent this recurrence. Efficacy is highly dependent on dose (55–70 Gy) (39,54,55). Doses of 45 to 60 Gy are associated with poor progression-free and overall survival rates. Recurrence rates from 50% to 100% have been reported in chordomas previously treated with conventional irradiation (52,53). The optimal dose for cranial chordomas is unknown. The Proton Radiation Oncology Group is studying doses of 75.6 and 82.9 CGE (56). The Proton Radiation Oncology Group also found excellent results in patients with skull base chondrosarcomas treated with 69.9 CGE (56). After mixed photon–proton beam irradiation of chondrosarcomas at 66 to 83 CGE, the 5- and 10year local recurrence-free survival rates were 97% and 92%, respectively (57). For chordomas, the 5- and 10-year local recurrence-free survival rates were 64% and 42%, respectively (57). The goal is to deliver these doses in a highly conformal way which spares adjacent radiosensitive structures from exposure beyond their tolerance. For example, exclusion of the optic nerve/chiasm at 55 to 60 CGE (58) may protect vision, and exclusion of the pituitary at 50 CGE may prevent endocrinopathies (59). Currently popular techniques include proton beam, intensity modulation radiotherapy, and stereotactic radiosurgery. Proton beam irradiation achieves high conformality by virtue of the Bragg peak effect of energy deposition by protons traversing tissue. The energy deposition profile of protons through tissue has a low entrance dose, a peak whose depth can be modulated, and no exit dose (60). This permits planning of treatments with high dose to tumor and rapid falloff in the surround. This high conformality is important for irregularly shaped tumors such as chordomas and chondrosarcomas at the skull base. In one series, 90% of the patients underwent pure proton beam treatment of chordomas with 65 to 79 CGE; they demonstrated a local control rate of 59% and an overall survival rate of 79% at 5 years (61). For chondrosarcomas, the 5-year local control rate and overall survival rate at 5 years are 75% and 100%, respectively (61). Particles other than protons, such as helium, neon, carbon, and neutrons, have also been used to treat chordomas and chondrosarcomas (62–66). Intensity modulation radiotherapy achieves the same goal by modulating the dose intensity pattern of the radiation to match the tumor’s shape. Treatment is carefully planned
500
Sakamoto and Harsh
using 3-D CT images and computerized dose calculations. Combinations of several intensity-modulated fields from different beam directions produce a custom tailored radiation dose that maximizes tumor dose while minimizing exposure of adjacent normal tissues. Stereotactic radiosurgery targets high doses of X-rays from a linear accelerator (LINAC or Cyberknife) or gamma rays from multiple cobalt sources (Gamma Knife) to a tumor. Numerous low-intensity beams from multiple directions intersect within the tumor volume so as to deliver high dose to the tumor while sparing surrounding tissue. Radiosurgery, traditionally, uses one to five treatment sessions. The precision possible permits treatment of tumor very close (within 2–3 mm) to critical structures (67). In that the risk of injury rises with targeted volume, radiosurgery is usually limited to tumor volumes less than 10 mL. Radiosurgery may be, particularly, useful to treat small unresectable residua of chondrosarcomas or recurrences of either chondrosarcomas or chordomas. All three of these modalities share the risk of neurologic deficit from injury to adjacent structures, such as delayed visual deterioration (2% in one proton beam series) (61), hypopituitarism (as high as 72% at 5 years and 84% at 10 years in one proton beam series and as low as 7% in another) (59,61), and hearing loss (10% in one proton beam series) (61). Use of lower doses may prevent these deficits. Use of radiation sensitizers such as razoxane may also improve the risk benefit calculus (68). Interstitial brachytherapy employs a different strategy. It involves permanent placement of radioactive seeds such as 125 I within a tumor or along a tumor resection margin. Gold foil implants are used to shield vital structures such as the brain stem from radiation. The half-life of 125 I is 60.2 days and almost all of the dose is delivered within four halflives. The dose to the margin is usually at least 50 Gy. The continuous low dose of radiation theoretically targets more cells as they move from radiation resistant to more radiation sensitive phases of the cell cycle. Additionally, the low dose should allow normal tissue to repair the damage caused by the sublethal doses of radiation.
Chemotherapy As therapies improve the local control of chordomas and chondrosarcomas, effective treatment of metastatic disease becomes increasingly important. Approximately 30% to 40% of patients with chordomas develop metastases (69,70), most commonly to the lungs, liver, and bone (71). Chemotherapy is an option for surgically inaccessible, previously irradiated recurrent local or metastatic tumor. Historically, it has had poor efficacy (44). Newer agents with tumor-specific rationales, such as imatinib mesylate, an inhibitor of tyrosine kinases activated by growth factors expressed in chordomas, may be an improvement (72).
OUTCOME AND PROGNOSIS Chordomas and chondrosarcomas, despite being lumped together historically, have distinct natural histories. The natural history of skull base chordomas is dismal. If not treated, the average affected patient will live approximately 18 months (6,73). Even with treatment, the 5-year survival rate ranges from 51% to 79% (13,18,22,61) and the 10-year survival rate ranges from 35% to 69% (18,22) (Fig. 5). Treatment usually involves attempted radical resection and radiotherapy. The importance of the extent of the initial resection is empha-
Figure 5 Local recurrence-free survival for skull base chordomas and chondrosarcomas after surgical resection (if necessary) and mixed proton– photon radiotherapy. Source: Reprinted with permission from Harsh G, ed., Chordomas and Chondrosarcomas of the Skull Base.
sized in one study which had a 5-year survival rate of 100% in patients with a radical or total excision without immediate postoperative radiotherapy (18). The overall operative mortality rate is approximately 1.9% to 5% (13,18,74). Subsequent operations carry greater risk to neurologic function and survival. A recent study demonstrated a 0% operative mortality rate for the first operation and a 7.1% mortality rate for the second (18). Surgeryrelated complications include cranial nerve palsies, stroke causing brain stem injury, CSF leakage, and meningitis. Permanent neurologic deficits were observed in 28.6% of patients in one series (74). Postoperative radiotherapy is almost always indicated. In one series, patients treated with surgery alone had a mean survival of 18 months versus 63 months for those who received postoperative radiotherapy (6). In patients with subtotal resections, postoperative conventional radiotherapy confers a 5-year survival rate of 65% (18). Patients who received postoperative proton beam therapy had an actuarial 5-year survival rate of 79% (61). Recurrences were felt to be due to limitations on delivered dose by critical structures. A larger series demonstrated 5-year and 10year local recurrence-free rates of 64% and 42%, respectively (57), as well as a complication rate of 8% (57). Although the series is small and the follow-up is limited, carbon ion therapy provided a 4-year survival rate of 86% (65). Initial results for patients who underwent subtotal resection followed by stereotactic radiosurgery appear promising: a 97 to 100% 2-year survival rate (75,76) and a 5-year survival rate of 82% (76). These series were small and the follow-up is limited. A more recent study shows actuarial tumor control rates at 2 years and 5 years of 93% and 52%, respectively (77). Recurrences are associated with a worse prognosis. The actuarial 3-year and 5-year survival rates for a chordoma, locally recurrent after surgery and radiotherapy, are 44% and 5%, respectively (78). Other factors associated with poor prognosis are tumor volume (>70cm3 ) (21,61) and older age of the patient (22). In one series, younger patients had 5-year and 10year survival rates of 75% and 63%, respectively, compared to 30% and 11%, respectively, in older patients (18).
Chapter 35: Chordoma and Chondrosarcoma of the Skull Base
Various immunohistochemical stains have been used as a prognostic indicator. A high MIB labeling index and overexpression of the tumor suppressor, p53, carry poor prognoses (79,80). Conventional chondrosarcomas have a better prognosis than do chordomas. Five-year survival rates depend on the histological grade of the tumor. Grade I, II, and III lesions had, respectively, 90%, 81%, and 43% 5-year survival rates in one series (37). Overall, the 5-year survival rate is 90% to 99% and the 10-year survival rate is 71% to 99% (13,15,81,82). Most surgical series advocate radical resection when feasible. Postoperative proton beam radiotherapy and radiosurgery are the most commonly used adjunctive treatments for small residual or recurrent tumors. Radical excision of low-grade chondrosarcomas may obviate the need for immediate postoperative irradiation. Among patients whose tumor was completely removed but who were not given postoperative radiotherapy, 78.3% experienced 5 years of recurrence-free survival (17,81). Morbidity and mortality rates for surgical resection of chondrosarcomas are similar to those for chordomas. Postoperative radiotherapy is indicated for most subtotally resected low-grade chondrosarcomas and all highergrade chondrosarcomas. It is highly effective. In a large series using proton beam radiotherapy, the 5-year and 10-year local recurrence-free survival rates were 97% and 92%, respectively, for chondrosarcomas (57). A smaller, more recent series demonstrated a 5-year survival rate of 100% (61). Radiosurgery is an attractive alternative for smaller residual tumors (82). In one series, postoperative fractionated stereotactic radiotherapy provided a 100% 5-year recurrence-free survival rate (76).
SUMMARY Chordomas and chondrosarcomas of the skull base are rare and challenging tumors. The relative inaccessibility of the skull base, the tumors’ proximity to critical neurovascular structures, their relative radioresistance to standard doses, their lack of chemosensitivity, and their tendency to recur locally require thoughtful choice and meticulous administration of therapy. Despite these challenges, the combination of surgical resection and high-dose radiation can cure almost all low-grade chondrosarcomas and provide meaningful intervals of control of both chordomas and high-grade chondrosarcomas.
REFERENCES 1. Luschka H. Die Altersveranderungen der Zwischenwirbelknorpel. Virchows Arch Path Anat. 1856;9:312–327. 2. Virchow R. Untersuchungen uber die entwickelung des schadelgrundes im gesunden und krankhaften zustande, und uber den einfluss derselben auf schadelform, gesichtsbildung und gehirnbau. G. Zeimer, 1857: Berlin. 3. Muller H. Ueber das Vorkommen von Resten der Chorda dorsalis bei Mecschen nach der Geburt und uber ihr Verhaltnis zu den Gallertgeschwulsten am Clivus. Ztschr Rat Med. 1858;2:202–229. 4. Klebs E. Ein Fall von Ecchondrosis spheno-occipitalise amylacea. Virchows Arch Path Anat. 1864;31:396–399. 5. Ribbert H. Ueber die Ecchondrosis physlifora sphenooccipitalis. Centralbl Allg Path Path Anat. 1894;5:457–461. 6. Heffelfinger MJ, et al. Chordomas and cartilaginous tumors at the skull base. Cancer. 1973;32(2):410–420.
501
7. Borba L, Al-Mefty O. Skull-base chordomas. Contemp Neurosurg. 1998;20:1–6. 8. Steenberghs J, et al. Intradural chordoma without bone involvement. Case report and review of the literature. J Neurosurg. 2002;97(1 Suppl):94–97. 9. Nishigaya K, et al. Intradural retroclival chordoma without bone involvement: No tumor regrowth 5 years after operation. Case report. J Neurosurg. 1998;88(4):764–768. 10. Laws E, Thapar K. Parasellar lesions other than pituitary adenomas. In: Powell M, Lightman SL, eds. Management of Pituitary Tumors: A Handbook. New York: Churchill-Livignstone, 1996:175–222. 11. Ewing J. A review of the classification of bone tumors. Surg Gynecol Obstet. 1939;68:971–976. 12. Neff B, et al. Chondrosarcoma of the skull base. Laryngoscope. 2002;112(1):134–139. 13. Gay E, et al. Chordomas and chondrosarcomas of the cranial base: Results and follow-up of 60 patients. Neurosurgery. 1995;36(5):887–896; discussion 896–897. 14. Hassounah M, et al. Primary cranial and intracranial chondrosarcoma. A survey. Acta Neurochir (Wien). 1985;78(3– 4):123–132. 15. Rosenberg AE, et al. Chondrosarcoma of the base of the skull: A clinicopathologic study of 200 cases with emphasis on its distinction from chordoma. Am J Surg Pathol. 1999;23(11):1370– 1378. 16. Koch BB, et al. National cancer database report on chondrosarcoma of the head and neck. Head Neck. 2000;22(4):408–425. 17. Crockard HA, et al. A multidisciplinary team approach to skull base chondrosarcomas. J Neurosurg. 2001;95(2):184–189. 18. Crockard HA, et al. A multidisciplinary team approach to skull base chordomas. J Neurosurg. 2001;95(2):175–183. 19. McMaster ML, et al. Chordoma: Incidence and survival patterns in the United States, 1973–1995. Cancer Causes Control. 2001;12(1):1–11. 20. Berkmen YM, Blatt ES. Cranial and intracranial cartilaginous tumours. Clin Radiol. 1968;19(3):327–333. 21. O’Connell JX, et al. Base of skull chordoma. A correlative study of histologic and clinical features of 62 cases. Cancer. 1994;74(8):2261–2267. 22. Forsyth PA, et al. Intracranial chordomas: A clinicopathological and prognostic study of 51 cases. J Neurosurg. 1993;78(5):741– 747. 23. Watkins L, et al. Skull base chordomas: A review of 38 patients, 1958–1988. Br J Neurosurg. 1993;7(3):241–248. 24. Wold LE, Laws ER Jr. Cranial chordomas in children and young adults. J Neurosurg. 1983;59(6):1043–1047. 25. Unni K. Dahlin’s Bone Tumors: General Aspects and Data on 11,087 Cases, 5th ed. Philadelphia, PA: Lippincott-Raven, 1996:291–303. 26. Yang XR, et al. Corroboration of a familial chordoma locus on chromosome 7q and evidence of genetic heterogeneity using single nucleotide polymorphisms (SNPs). Int J Cancer. 2005;116(3):487–491. 27. Laws E Jr. Cranial Chordomas. In: Wilkins R, Rengachary S, eds. Neurosurgery. New York: McGraw-Hill, 1985:927–929. 28. Thodou E, et al. Intrasellar chordomas mimicking pituitary adenoma. J Neurosurg. 2000;92(6):976–982. 29. Campbell WM, et al. Nasal and paranasal presentations of chordomas. Laryngoscope. 1980;90(4):612–618. 30. Cianfriglia F, Pompili A, Occhipinti E. Intracranial malignant cartilaginous tumours. Report of two cases and review of literature. Acta Neurochir (Wien). 1978;45(1–2):163–175. 31. Oguro K, et al. Chondrosarcoma of the posterior fossa—case report. Neurol Med Chir (Tokyo). 1989;29(11):1030–1038. 32. Korten AG, et al. Intracranial chondrosarcoma: Review of the literature and report of 15 cases. J Neurol Neurosurg Psychiatry. 1998;65(1):88–92. 33. Mandahl N, et al. Cytogenetic aberrations and their prognostic impact in chondrosarcoma. Genes Chromosomes Cancer. 2002;33(2):188–200. 34. Tachibana E, et al. Surgical treatment of a massive chondrosarcoma in the skull base associated with Maffucci’s syndrome:
502
35. 36. 37.
38. 39.
40.
41. 42. 43. 44. 45.
46. 47.
48. 49. 50.
51.
52. 53.
54. 55.
56.
57. 58. 59.
Sakamoto and Harsh A case report. Surg Neurol. 2000;54(2):165–169; discussion 169– 170. Ariel IM, Verdu C. Chordoma: An analysis of twenty cases treated over a twenty-year span. J Surg Oncol. 1975;7(1):27–44. Richardson MS. Pathology of skull base tumors. Otolaryngol Clin North Am. 2001;34(6):1025–1042, vii. Evans HL, Ayala AG, Romsdahl MM. Prognostic factors in chondrosarcoma of bone: A clinicopathologic analysis with emphasis on histologic grading. Cancer. 1977;40(2):818–831. Finn DG, Goepfert H, Batsakis JG. Chondrosarcoma of the head and neck. Laryngoscope. 1984;94(12 Pt 1):1539–1544. Raffel C, et al. Cranial chordomas: Clinical presentation and results of operative and radiation therapy in twenty-six patients. Neurosurgery. 1985;17(5):703–710. Matsumoto J, Towbin RB, Ball WS Jr. Cranial chordomas in infancy and childhood. A report of two cases and review of the literature. Pediatr Radiol. 1989;20(1–2):28–32. Menezes AH, Traynelis VC, Gantz BJ. Surgical approaches to the craniovertebral junction. Clin Neurosurg. 1994;41:187–203. Meyer FB, Ebersold MJ, Reese DF. Benign tumors of the foramen magnum. J Neurosurg. 1984;61(1):136–142. Levy ML, Chen TC, Weiss MH. Monostotic fibrous dysplasia of the clivus. Case report. J Neurosurg. 1991;75(5):800–803. Castellano F, Ruggiero G. Meningiomas of the posterior fossa. Acta Radiol Suppl. 1953;104:1–177. Brisman JL, et al. Eosinophilic granuloma of the clivus: Case report, follow-up of two previously reported cases, and review of the literature on cranial base eosinophilic granuloma. Neurosurgery. 1997;41(1):273–278; discussion 278–279. Meyer JE, Oot RF, Lindfors KK. CT appearance of clival chordomas. J Comput Assist Tomogr. 1986;10(1):34–38. Grossman RI, Davis KR. Cranial computed tomographic appearance of chondrosarcoma of the base of the skull. Radiology. 1981;141(2):403–408. Brown E, Hug EB, Weber AL. Chondrosarcoma of the skull base. Neuroimaging Clin N Am. 1994;4(3):529–541. Al-Mefty O, Borba LA. Skull base chordomas: A management challenge. J Neurosurg. 1997;86(2):182–189. Frank G, et al. The endoscopic transnasal transsphenoidal approach for the treatment of cranial base chordomas and chondrosarcomas. Neurosurgery. 2006;59(1 Suppl 1):ONS50–7; discussion ONS50–7. Sen CN, Sekhar LN. An extreme lateral approach to intradural lesions of the cervical spine and foramen magnum. Neurosurgery. 1990;27(2):197–204. Catton C, et al. Chordoma: Long-term follow-up after radical photon irradiation. Radiother Oncol. 1996;41(1):67–72. Cummings BJ, Hodson DI, Bush RS. Chordoma: The results of megavoltage radiation therapy. Int J Radiat Oncol Biol Phys. 1983;9(5):633–642. Higinbotham NL, et al. Chordoma. Thirty-five-year study at Memorial Hospital. Cancer. 1967;20(11):1841–1850. Pearlman AW, Friedman M. Radical radiation therapy of chordoma. Am J Roentgenol Radium Ther Nucl Med. 1970;108(2):332–341. Liebsch NJ, Munzenrider JE. Proton Radiotherapy for Cranial Base Chordomas. In: Harsh G, ed. Chordomas and Chondrosarcomas of the Skull Base and Spine. New York: Thieme, 2003:307– 314. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull base. Strahlenther Onkol. 1999;175(Suppl 2):57–63. Habrand IL, et al. Neurovisual outcome following proton radiation therapy. Int J Radiat Oncol Biol Phys. 1989;16(6):1601–1606. Pai HH, et al. Hypothalamic/pituitary function following highdose conformal radiotherapy to the base of skull: Demonstration of a dose-effect relationship using dose-volume histogram analysis. Int J Radiat Oncol Biol Phys. 2001;49(4):1079–1092.
60. Hug EB. Review of skull base chordomas: Prognostic factors and long-term results of proton-beam radiotherapy. Neurosurg Focus. 2001;10(3):E11. 61. Hug EB, et al. Proton radiation therapy for chordomas and chondrosarcomas of the skull base. J Neurosurg. 1999;91(3):432– 439. 62. Berson AM, et al. Charged particle irradiation of chordoma and chondrosarcoma of the base of skull and cervical spine: The Lawrence Berkeley Laboratory experience. Int J Radiat Oncol Biol Phys. 1988;15(3):559–565. 63. Castro JR, et al. Experience in charged particle irradiation of tumors of the skull base: 1977–1992. Int J Radiat Oncol Biol Phys. 1994;29(4):647–655. 64. Saunders WM, et al. Precision, high dose radiotherapy. II. Helium ion treatment of tumors adjacent to critical central nervous system structures. Int J Radiat Oncol Biol Phys. 1985;11(7):1339– 1347. 65. Schulz-Ertner D, et al. Carbon ion radiation therapy for chordomas and low grade chondrosarcomas—current status of the clinical trials at GSI. Radiother Oncol. 2004;73(Suppl 2): S53–S56. 66. Schulz-Ertner D, et al. Results of carbon ion radiotherapy in 152 patients. Int J Radiat Oncol Biol Phys. 2004;58(2):631–640. 67. Muthukumar N, et al. Stereotactic radiosurgery for chordoma and chondrosarcoma: Further experiences. Int J Radiat Oncol Biol Phys. 1998;41(2):387–392. 68. Rhomberg W, et al. Combined radiotherapy and razoxane in the treatment of chondrosarcomas and chordomas. Anticancer Res. 2006;26(3B):2407–2411. 69. Chambers PW, Schwinn CP. Chordoma. A clinicopathologic study of metastasis. Am J Clin Pathol. 1979;72(5):765–776. 70. Sundaresan N, et al. Spinal chordomas. J Neurosurg. 1979;50(3):312–319. 71. O’Neill P, et al. Fifty years of experience with chordomas in southeast Scotland. Neurosurgery. 1985;16(2):166–170. 72. Casali PG, et al. Imatinib mesylate in chordoma. Cancer. 2004;101(9):2086–2097. 73. Eriksson B, Gunterberg B, Kindblom LG. Chordoma. A clinicopathologic and prognostic study of a Swedish national series. Acta Orthop Scand. 1981;52(1):49–58. 74. Colli BO, Al-Mefty O. Chordomas of the skull base: Follow-up review and prognostic factors. Neurosurg Focus. 2001;10(3):E1. 75. Miller RC, et al. The role of stereotactic radiosurgery in the treatment of malignant skull base tumors. Int J Radiat Oncol Biol Phys. 1997;39(5):977–981. 76. Debus J, et al. Stereotactic fractionated radiotherapy for chordomas and chondrosarcomas of the skull base. Int J Radiat Oncol Biol Phys. 2000;47(3):591–596. 77. Krishnan S, et al. Radiosurgery for cranial base chordomas and chondrosarcomas. Neurosurgery. 2005;56(4):777–784; discussion 777–784. 78. Fagundes MA, et al. Radiation therapy for chordomas of the base of skull and cervical spine: Patterns of failure and outcome after relapse. Int J Radiat Oncol Biol Phys. 1995;33(3):579– 584. 79. Naka T, et al. Alterations of G1-S checkpoint in chordoma: The prognostic impact of p53 overexpression. Cancer. 2005;104(6):1255–1263. 80. Saad AG, Collins MH. Prognostic value of MIB-1, E-cadherin, and CD44 in pediatric chordomas. Pediatr Dev Pathol. 2005;8(3):362–368. 81. Tzortzidis F, et al. Patient outcome at long-term follow-up after aggressive microsurgical resection of cranial base chondrosarcomas. Neurosurgery. 2006;58(6):1090–1098; discussion 1090– 1098. 82. Wanebo JE, et al. Management of cranial base chondrosarcomas. Neurosurgery. 2006;58(2):249–255; discussion 249–255.
36 Meningioma Ashwin Viswanathan and Franco DeMonte
INCIDENCE/EPIDEMIOLOGY Incidence
the most common radiation-induced neoplasm (14). Studies of immigrants to Israel who had received radiotherapy for tinea capitis between 1948 and 1960 clearly demonstrate therapeutic radiation as an etiological factor (15). Patients who had received radiation showed a relative risk of 9.5 for the development of meningioma, oftentimes after a latency period of 20 to 40 years. Similarly, studies of atomic bomb survivors in Hiroshima and Nagasaki have shown a relative risk of 6.48 when compared with non-exposed populations (16). Both increasing time from exposure and increasing dose of radiation are associated with a higher incidence of meningiomas (17). As compared with non-radiation–induced meningiomas, radiation-induced meningiomas tend to possess atypical histological features, have a more aggressive clinical course, occur at multiple locations, and have different cytogenetic characteristics (14). Cellular phone usage and the concomitant exposure to low-dose radiation exposure has recently generated attention as a possible etiological factor in the formation of brain tumors. However, to date no conclusive evidence has shown a link between cell phone usage and the development of meningiomas (18,19).
Meningiomas account for 32% of all primary brain tumors and represent the second most common central nervous system (CNS) neoplasm in adults after gliomas (1). The incidence of meningioma is 4.7 cases per 100,000 person-years, and women outnumber men by 2.1:1. The median age at diagnosis is 63 and the incidence of meningioma increases with increasing age. The incidence of diagnostically and nondiagnostically confirmed meningioma has increased between 1985 and 1999, with the incidence of nondiagnostically confirmed meningioma increasing by 4.1% per year. This trend likely reflects the increased use of MRI in the diagnosis of incidental meningioma (2). Data from the population-based Rotterdam study found incidental meningiomas in 18 out of 2000 (0.9%) MRI scans performed. These incidental meningiomas ranged from 5 mm to 60 mm in diameter and the prevalence was 1.1% in women and 0.7% in men (3). Intracranial meningiomas outnumber spinal meningiomas by approximately 10:1 (4). Pediatric meningiomas are rare and comprise less than 2% of all meningiomas and less than 5% of all pediatric brain tumors (5,6). The most common locations for intracranial meningiomas are parasagittal, sphenoid ridge, and convexity. Forty percent of all meningiomas arise from the base of the anterior, middle, or posterior fossa and are the most common skull base tumors. Sphenoid wing meningiomas make up about half of these; tuberculum sella and olfactory groove tumors the other half. Ectopic meningiomas have been described in the orbit, paranasal sinuses, skin, subcutaneous tissues, lung, mediastinum, and adrenal glands. Table 1 details the common sites for meningiomas and their incidence.
Infection The role of infectious agents in the development of brain tumors has also been investigated. The number of siblings in a family has been proposed as an indicator for exposure to infectious agents (more siblings relates to a greater risk or exposure to infectious agents). For patients less than 15 years of age at the time of meningioma diagnosis, a rate ratio of 3.71 was found when compared with patients with no siblings (20). Viruses, and in particular Simian virus 40 (SV40), have been studied as an etiological agent as well. SV40, a polyomavirus, is capable of transforming cells into those with a neoplastic phenotype (21). The oncogenic and transforming properties of SV40 are related to expression of large tumor antigen (Tag), which is postulated to have a role in inactivating the tumor suppressor functions of p53, pRb, p107, and others (22). Of 10 human meningioma samples analyzed, SV40 Tag was identified in 7 samples, the Tag-p53 complex was identified in 3, and the Tag-pRb complex in 2 (23). Rollison et al., however, were not able to identify SV40 in any of the 15 human meningioma tumor samples they analyzed (24). Further work is necessary to delineate and characterize a potential relationship.
Etiology Trauma Several case-control studies have shown an elevated odds ratio for the development of meningioma in patients with a history of head trauma, though no clear association has been found to date (7–10). A recent population-based case-control study of 200 patients with meningioma found an elevated odds ratio of 2.62 for the development of meningioma given a history of head trauma (11). In patients with a history of head trauma occurring 10 to 19 years prior to the date of diagnosis, an odds ratio of 4.33 was found. Yet other studies have found no such elevated risk (12,13). Recall bias has been suggested as a confounding factor limiting the effectiveness of case-control studies. Further epidemiological studies are necessary to validate and delineate a relationship.
Genetics NF-2 Mutations The majority of meningiomas are sporadic tumors in patients with no history of brain tumors. However, familial meningiomas have been identified in a number of conditions including neurofibromatosis type 2 (NF2), NF1, Gorlin/nevoid basal cell carcinoma syndrome, Rubinstein–Taybi
Radiation Exposure to ionizing radiation is a known etiological factor in the development of meningiomas, and meningiomas are 503
504
Viswanathan and DeMonte Table 1
Location and Incidence of Intracranial Meningioma
Site Parasagittal/Falcine Convexity Sphenoid ridge Suprasellar (tuberculum) Posterior fossa Olfactory groove Middle fossa/Meckel cave Tentorial Peritorcular Lateral ventricle Foramen magnum Orbit/optic nerve sheath
Incidence (%) 25 19 17 9 8 8 4 3 3 1–2 1–2 1–2
Source: Based on data from Refs. 55, 91–94.
syndrome, Li–Fraumeni syndrome, von Hippel–Lindau syndrome, Werner syndrome, Gardner syndrome, and the melanoma/astrocytoma-brain tumor syndrome (25). Meningiomas were one of the first solid tumors to be associated with a characteristic cytogenetic change—loss of heterozygosity (LOH) of chromosome 22. Forty to seventy percent of meningiomas exhibit loss of heterozygosity (LOH) for markers from the chromosomal region 22q12.2 that encompasses the NF2 gene. Mutations specifically in the NF2 gene have been reported in 30% to 60% of sporadic meningiomas. The frequency of NF2 mutations is similar among WHO grades I, II, and III meningiomas (26). This finding suggests that NF2 gene inactivation may be an important initiation step in the formation of meningiomas, but may not play a role in tumor progression (25). Inactivation of both NF2 gene alleles is necessary for tumor formation. The NF2 protein merlin belongs to the protein 4.1 families that link membrane proteins to the cytoskeleton. Hannson et al. utilized microarray-based comparative genomic hybridization to study 126 sporadic meningioma specimens (27). They found the incidence of biallelic NF2 inactivation to be 52% in fibroblastic variants as compared with 18% in meningothelial histologies. This finding suggests that NF2 inactivation may not be a critical step in the formation of meningothelial meningiomas (27). In addition to NF2 mutations, other protein 4.1 family members are also downregulated in meningiomas. The EPB41L3 or DAL1 gene product protein 4.1B belongs to the same superfamily as merlin, and is thought to act as a tumor suppressor. Loss of protein 4.1B expression is common in meningiomas of all grades, though one group found the loss predominantly in meningiomas of higher grade (28). However, no propensity to develop tumors has been observed in transgenic mice lacking EPB41L3/DAL1 (29).
Other Genetic Abnormalities Deletion of 1p is the second most common genetic change seen in meningiomas (27), and has been associated with meningioma progression and recurrence. Co-deletion of 1p and 22q is seen in 67% of meningiomas. Transition to atypical meningioma has been associated with losses on chromosomes 1p, 6q, 10, 14q, and 18q and gains on chromosomes 1q, 9q, 12q, 15q, 17q, and 20q (30). Grade III or anaplastic tumors are associated with gains on 17q23 and losses on 9p. They may also demonstrate more frequent losses on 6q, 10, and 14q as compared with atypical tumors. Loss of chromosome 14 represents the third most common genetic abnormality after mutations in chromosomes 22 and 1. Deletions on chromosome 14 have been associated
with an increased risk of relapse and poorer prognosis (31). More recent DNA microarray assays have shown losses on chromosomes 10 and 14 in high-grade meningiomas with increased expression of several genes related to IGF (IGF2, IGFBP3, AKT3) or wingless (WNT, CTNNB1, CDK5R1, ENC1, CCND1) pathways (32). Other studies have demonstrated amplification of MSH2 (2p22.3-p22.1) in 16 of 31 meningiomas (51.6%), deletion of GSCL (22q11.21) in 41.9%, amplification of INS (12ptel) and TCL1A (14q32.1), and deletions of HIRA (22q11.21) and IGH (14qtel) (33). Recently, Pecina-Slaus et al. showed loss of heterozygosity of the Adenomatous polyposis coli (APC) gene in 47% of 32 specimens (34). APC acts as a tumor suppressor gene and has previously been recognized as a colon-specific tumor suppressor.
Tumor Biology Growth Factors Meningiomas have been found to express a number of growth factors and their receptors including epidermal growth factor receptor (EGFR), basic fibroblast growth factor (BFGF) receptor, platelet-derived growth factor (PDGF) receptor, and vascular endothelial growth factor-A (VEGF-A). This finding has led to the possibility that autocrine growth factor secretion and autocrine loops may play a role in the growth of meningiomas (25). Recent investigation by Smith et al. demonstrated PDGFR-β expression in all of the 84 meningioma samples studied (35). In addition, expression of BFGFR was found in 89% of benign meningiomas, while EGFR immunoreactivity was detected in 47% of benign meningiomas. EGFR immunoreactivity was found as a strong predictor of prolonged survival in patients with atypical meningioma (35). VEGF-A, which is also known as vascular permeability factor, is considered to be a key factor in angiogenesis and edema formation for meningiomas. Several studies have demonstrated VEGF-A levels in meningiomas to be associated with the extent of peritumoral edema (36,37) and some smaller studies have postulated that VEGF-A mRNA expression may correlate with meningioma vascularity (37,38). Other studies however have found no association between VEGF-A protein levels and microvessel density (39).
Receptors The increased incidence of meningioma in women, along with early studies reporting accelerated tumor growth during pregnancy (40) and associations with breast cancer (41), has led to investigation of the role of sex steroids in the pathogenesis of meningioma. Most studies have found meningiomas to lack estrogen receptors (ERs), though some have found low concentrations of ERs in 5% to 33% of meningiomas (42,43). Progesterone receptors are much more common, being expressed in about two-thirds of meningiomas. As seen in breast cancer, expression of the progesterone receptor tends to decrease during malignant progression (25,44). Lusis et al. utilized high-throughput tissue microarray immunohistochemistry (TMA-IHC) to study 41 meningioma samples of all histological grades. They found PR reactivity in all benign specimens, as compared with 67% of atypical meningiomas, and 56% of anaplastic meningiomas (44). Pravdenkova et al. found the expression of the PR alone indicates a more favorable clinical and biological outcome (45).
PATHOLOGY Meningiomas arise from the arachnoidal (meningothelial) cells, and most are well-demarcated tumors with a broad
Chapter 36: Meningioma
505
Table 2 World Health Organization 2007 Classification of Meningiomas Meningiomas with low risk of recurrence and aggressive growth Meningothelial meningioma WHO grade I Fibrous (fibroblastic) meningioma WHO grade I Transitional (mixed) meningioma WHO grade I Psammomatous meningioma WHO grade I Angiomatous meningioma WHO grade I Microcystic meningioma WHO grade I Secretory meningioma WHO grade I Lymphoplasmacyte-rich meningioma WHO grade I Metaplastic meningioma WHO grade I Meningiomas with greater likelihood of recurrence and/or aggressive behavior Chordoid meningioma WHO grade II Clear cell meningioma (intracranial) WHO grade II Atypical meningioma WHO grade II Papillary meningioma WHO grade III Rhabdoid meningioma WHO grade III Anaplastic (malignant) meningioma WHO grade III Meningiomas of any subtype or grade with high proliferation index and/or brain invasion
dural attachment. The World Health Organization (WHO) 2007 classification divides meningiomas into three grades: benign (WHO grade I), atypical (WHO grade II), and anaplastic or malignant (WHO grade III) (46). Approximately 80% of meningiomas are WHO grade I and possess a low risk of recurrence or aggressive growth. Grade II meningiomas account for 15% to 20% of all meningiomas, while grade III tumors comprise only 1% to 3% meningiomas (31). Grade II and grade III meningiomas are characterized by a greater likelihood of recurrence or of aggressive growth. Table 2 groups the subtypes of meningiomas based on their likelihood of recurrence. The WHO 2007 classification is based upon the number of mitotic figures (MFs) per 10 high-powered fields (HPF) in the area of highest mitotic activity. Meningiomas of any grade may exhibit invasion of brain parenchyma, which is characterized by finger-like projections of tumor cells without an intervening layer of leptomeninges. The presence of brain invasion does confer a greater risk for recurrence similar to that seen for atypical meningiomas. Proliferative indices are not currently included in the grading criteria for meningiomas due to significant difference in technique and interpretation between laboratories (46).
WHO Grade I Meningothelial, fibrous, and transitional meningiomas are the three most common histological variants of grade I meningiomas. Less common subtypes include psammomatous, angiomatous, microcystic, and secretory meningiomas. Chordoid, clear-cell, papillary, and rhabdoid variants of meningiomas are often associated with more aggressive tumor behavior and are consequently not classified as grade I tumors. Benign meningiomas may also invade surrounding structures including the skull, dural sinuses, orbit, and soft tissues. Meningothelial meningioma is a common variant in which tumor cells form lobules surrounded by thin collagenous septae (Fig. 1). Tumor cells are uniform, resembling normal arachnoid cap cells, and may show central clearing. Rounded eosinophilic cytoplasmic protrusions termed pseudoinclusions are also seen. Whorls and psammoma bodies, when present, are less well defined than those seen in fibrous, psammomatous, or transitional meningiomas. Fibrous meningiomas are formed by spindle-shaped cells that resem-
Figure 1 WHO grade I meningioma with multiple “whorls” (arrows, H&E x200). Source: Property of the Department of Neurosurgery, The University of Texas M.D. Anderson Cancer Center—Used with permission.
ble fibroblasts. These spindle-shaped cells form intersecting fascicles and are embedded in a collagen-rich and reticulinrich matrix. Transitional meningioma (mixed meningioma) is a common histological subtype which contains features of both meningothelial and fibrous meningiomas. Both the lobular arrangement of meningothelial meningiomas and the fascicular pattern of fibrous variants may be seen next to one another. They usually demonstrate extensive whorl formation in which the tumor cells form concentric cell layers by wrapping around each other. When the whorl formations hyalinize and calcify, they form structures known as psammoma bodies.
WHO Grade II By the WHO 2007 criteria, a meningioma is classified as atypical if it has ≥4 mitoses per 10 high-power fields, or meets three of the five following criteria: increased cellularity, high nuclear to cytoplasmic ratio (small cells), prominent nucleoli, uninterrupted patternless or sheet-like growth, or foci of spontaneous (not induced by embolism) necrosis (46) (Fig. 2). Two meningioma variants, clear-cell and chordoid, have been found to have higher recurrence rates even in the absence of the above criteria, leading to their classification as atypical meningiomas. Clear cell meningioma is a rare meningioma variant which is composed of sheets of polygonal cells with clear, glycogen-rich cytoplasm that is positive for periodic acid Schiff. They also demonstrate dense perivascular and interstitial collagen deposition. Clear cell tumors are more commonly found in the cauda equina or the cerebellopointine angle, and have a tendency to affect younger patients. Chordoid meningiomas are typically supratentorial tumors and have regions that are histologically similar to chordoma. They may have cords of small epithelioid tumor cells that contain eosinophilic cytoplasm residing in a basophilic, mucinrich matrix (30).
WHO Grade III Malignant meningiomas are characterized by a highly elevated mitotic index of >20 mitoses per 10 high-power
506
Viswanathan and DeMonte
Immunohistochemistry
Figure 2 Atypical meningioma, WHO grade II. This photomicrograph illustrates cellular sheeting, a feature indicative of increased tumor aggressiveness (H&E x200). Source: Property of the Department of Neurosurgery, The University of Texas M.D. Anderson Cancer Center—Used with permission.
fields and can pathologically resemble sarcoma, carcinoma, or melanoma (Fig. 3). In addition, large areas of necrosis can usually be seen. Papillary and rhabdoid meningiomas are consistently associated with malignant behavior and are therefore classified as WHO grade III tumors (47). Papillary meningioma has a propensity to affect younger patients, commonly exhibits brain invasion, and may show distant metastases. The rhabdoid variant is a rare tumor which contains sheets of large rhabdoid cells with eosinophilic cytoplasm containing whorled intermediate filaments and eccentric nuclei. The median survival is less than 2 years and the recurrence rate is 50% to 80% after surgical resection (48).
Figure 3 Anaplastic meningioma, WHO grade III. Mitoses are easily identified in this anaplastic meningioma (arrows) as is the marked nuclear pleomorphism (H&E x200). Source: Property of the Department of Neurosurgery, The University of Texas M.D. Anderson Cancer Center—Used with permission.
Epithelial membrane antigen (EMA) is the most commonly used marker in pathological analysis of meningiomas and most tumors show at least scattered positivity for this antigen (49). All meningiomas also strongly express vimentin (50). Tissue microarray immunohistochemistry (TMA-IHC) has proven to be an efficient and reliable method for analyzing biomarkers in meningioma. Using TMA-IHC, Lusis et al. found EMA reactivity in 100% of meningiomas regardless of grade and E-cadherin immunoreactivity in 91% of all meningiomas and 90% of anaplastic meningiomas (44). However, these markers are not specific for meningiomas. Immunohistochemistry has also proved useful in quantifying the proliferative index for meningiomas with the antibody MIB-1, which targets the proliferation marker Ki-67. Elevated proliferative indices as measured by MIB-1 labeling have been associated with an increased risk of recurrence (51). Mean MIB-1 labeling indices are 0.7% to 2.2% for grade I meningiomas, 2.1% to 9.3% for grade II meningiomas, and 11% to 16.3% for grade III meningiomas (52). MIB-1 labeling indices of greater than 5% suggest a greater likelihood of recurrence. Similarly, an antibody that specifically recognizes the phosphorylated histone H3 (PHH3) has been effectively used as an aid in grading meningiomas (53). During mitosis, phosphorylation of the Ser-10 residue of histone H3 reaches a maximum. Consequently, immunostaining with anti-PHH3 allows the observer to rapidly focus on the most mitotically active areas of the tumor (Fig. 4).
TREATMENT Surgery Complete surgical excision is the treatment of choice for meningiomas. The ability to achieve a safe, complete resection is influenced by tumor involvement of major dural sinuses, association with eloquent neurovascular structures, tumor location and size, and previous surgery or radiation (54). The vast majority of meningiomas are surrounded by a layer of arachnoid which separates the tumor from the brain, cranial nerves, and blood vessels. Meningiomas may attach to or
Figure 4 pHH3 immunostaining readily identifies the presence of mitoses (x200). Source: Property of the Department of Neurosurgery, The University of Texas M.D. Anderson Cancer Center—Used with permission.
Chapter 36: Meningioma
surround cerebral arteries, but only very rarely do they invade the arterial walls. By accessing this arachnoidal plane, the surgeon is able to minimize the chance of injury to neurovascular structures. Internal debulking of the tumor facilitates the delineation of the arachnoid plane by allowing the edge of the tumor to collapse inward.
Convexity Meningiomas Convexity meningiomas comprise approximately 15% of all meningiomas and they possess the greatest potential for cure. By definition, convexity meningiomas do not arise from the skull base and do not involve the dural sinuses, and hence they allow for excision of a wide dural margin. Recurrent and/or aggressive meningiomas may not have the normal arachnoidal layer separating them from the cerebrum and require careful sharp dissection under the operating microscope to minimize cortical injury. Once the tumor and a wide dural margin have been resected, the dural defect may be repaired with pericranium, fascia lata, temporalis fascia, cadaveric dura, synthetic collagen matrix, or bovine pericardium.
Parasagittal Meningiomas Cushing and Eisenhardt defined the parasagittal meningioma as one that fills the parasagittal angle, with no brain tissue between the tumor and the superior sagittal sinus (55). Parasagittal meningiomas account for 17% to 32% of meningiomas and the primary consideration in their removal is management of the superior sagittal sinus and the cerebral veins that drain into it. Surgical approaches and management include simple dissection of the meningiomas off the lateral wall of the sinus, sagittal sinus reconstruction, and excision of the sinus in the case of a totally occluded sinus.
Olfactory Groove and Tuberculum Sellae Meningiomas Olfactory groove and tuberculum sellae tumors each comprise approximately 10% of meningiomas. Both tumors are midline lesions that derive their blood supply from the ethmoidal branches of the ophthalmic arteries, the anterior branch of the middle meningeal artery, and the meningeal branches of the ICA. Early division of this vascular supply is the first step in tumor removal. A low basal approach is preferred with a unilateral supraorbital craniotomy usually being sufficient. The optic nerves and chiasm are typically displaced and fine microdissection is required to free these structures. Inspection of both optic canals is necessary to detect tumor extension into this area—a common occurrence that can easily be missed.
Sphenoid Wing Meningiomas Sphenoid wing meningiomas are the second most common type of meningiomas after the parasagittal type. These meningiomas are classified according to their point of origin along the sphenoid ridge and include the meningiomas en plaque that are characterized by hyperostosis of the sphenoid bone that causes progressive painless proptosis and occasionally cranial neuropathies secondary to foraminal encroachment. The internal carotid, the middle and anterior cerebral arteries and their branches, as well as the optic, oculomotor, and olfactory nerves are the neurovascular structures at greatest risk during the surgical removal of sphenoid wing meningiomas. The presence of the arachnoidal layer allows for meningiomas to be microsurgically separated from these structures even though there may be a marked distortion of the normal anatomy of the region.
507
Posterior Fossa Meningiomas Posterior fossa meningiomas account for 10% of all intracranial meningiomas. Almost half of these meningiomas are located in the cerebellopontine angle, 40% are tentorial or cerebellar convexity tumors, and 9% and 6% are petroclival or at the foramen magnum. A standard retrosigmoid craniotomy allows sufficient exposure for the removal of most meningiomas of the cerebellopontine angle while supra- and infratentorial and presigmoid approaches may be necessary for petroclival meningiomas. Foramen magnum meningiomas may require a far lateral or trans-occipital condylar approach for optimal access. Following tumor debulking and devascularization the tumor is dissected from the brain stem the basilar, vertebral and cerebellar arteries and the trochlear, trigeminal, abducens, facial, vestibulocochlear, and lower cranial nerves. Meningiomas involving the tentorium and cerebellar convexity have the transverse sinus as their main area of concern. Management of the transverse sinus is similar to that of the sigmoid sinus in parasagittal meningiomas described above.
Radiation External Beam Radiation Therapy Though complete surgical resection is the ultimate goal in treating patients with meningioma, this is not always possible with an acceptable level of morbidity. Specifically, meningiomas of the sphenoid wing, cavernous sinus, clivus, cerebellopontine angle, and sellar regions are more likely to be subtotally excised (56). In particular, external beam radiation therapy (EBRT) has become an integral part of the management of optic nerve sheath meningiomas. In their evaluation of 64 patients with long-term follow-up, Turbin et al. concluded that EBRT led to more favorable outcomes as compared with surgical resection, observation, or surgical resection plus EBRT (57). Several other studies have shown similar results (58). Radiation therapy has evolved as an additional means for controlling meningioma in patients who have either undergone a subtotal resection or in patients with atypical or anaplastic histologies (59). Recommended doses generally range from 50 to 55 Gy in fractions of 1.8 to 2.0 Gy. The planning target volume can include only the gross tumor volume or the gross tumor volume plus a margin depending on the grade of the meningioma (2 cm margin recommended for anaplastic meningioma). Targeting the dural tail of meningiomas remains a subject of controversy (58).
Stereotactic Radiosurgery and Radiotherapy Stereotactic radiation techniques have emerged over the past 20 years as an important alternative to conventional external beam radiation therapy. Three modalities exist for stereotactic radiosurgery (SRS): LINAC, Gamma Knife, and protons. Tumors most appropriate for SRS are those that are smaller than 3.5 cm with little surrounding edema, and in locations where dose constraints for critical structures including the optic apparatus and the brainstem can be respected (60). Khoo et al. compared the clinical target volumes using CT and MRI for patients with skull base meningiomas undergoing radiation therapy. They found the MR and CT based target volumes provided complementary data regarding tumor involvement in soft tissue and bony regions respectively (61). Consequently, MR and CT fusion images are optimal for treatment planning of smaller meningiomas. For larger meningiomas, CT-based planning is usually adequate. Fractionated stereotactic radiotherapy (SRT) allows for precise stereotactic targeting, steep dose gradients, and the
508
Viswanathan and DeMonte
benefit of allowing normal tissues to heal between radiation fractions. LINAC is the primary modality used for SRT, and is used with a relocatable frame. Immobilization systems used with SRT include bite block devices such as the Gill–Thomas– Cosman frame for adults and the Tarbell–Loeffler–Cosman frame for children. After irradiation of benign meningiomas, surveillance MR images should be obtained at 6 and 12 months, and then annually thereafter.
Table 3 Simpson’s Classification for the Extent of Resection of Intracranial Meningioma
Chemotherapy
Grade IV Grade V
Chemotherapy has a very limited role in the treatment of meningiomas. No clear evidence of efficacy has been shown for chemotherapeutic agents, though a number of different agents have been investigated. Hydroxyurea is an oral ribonucleotide reductase inhibitor which arrests meningioma cell division in the S-phase of the cell cycle and induces apoptosis (62). Though this agent has been effective in in vitro and in vivo studies, treatment of patients with recurrent or unresectable meningiomas has shown little benefit (63–65). Combination of hydroxyurea with calcium channel blockers has also been investigated in vitro with positive results (65). In vitro studies of the combination of interferon-alpha and 5fluorouracil have been promising (66), and interferon-alpha has been shown effective in prolonging the time to recurrence in a small group of patients with aggressive meningioma (67). Temozolamide, an alkylating agent which has been used in malignant gliomas, showed no benefit in treating refractory meningiomas in a phase II trial (68). Mifepristone (RU486) is a progesterone blocker that has been shown to inhibit the growth of cultured human meningioma tissue and in animal models (69). Long-term administration of mifepristone showed only mild clinical benefit in 8 of 28 patients. Endometrial hyperplasia was noted in several patients after long-term administration (70). Irinotecan (CPT11), a topoisomerase-1 inhibitor, has been shown to inhibit in-vitro cultures of human meningioma cell lines and invivo studies using a subcutaneous tumor model (71). However, a phase II study evaluating CPT-11 in patients with recurrent meningioma was stopped prematurely as all patients demonstrated tumor progression within 6 months (72).
Grade I Grade II Grade III
Gross total resection of tumor, dural attachments, and abnormal bone Gross total resection of tumor, coagulation of dural attachments Gross total resection of tumor, without resection or coagulation of dural attachments, or alternatively of its extradural extensions (e.g., invaded sinus or hyperostotic bone) Partial resection of tumor Simple decompression (biopsy)
ing (73). Careful observation with an initial imaging study 3 months following the first is recommended to identify atypical or anaplastic growth patterns; a second scan 6 months later to detect any growth; and then yearly thereafter is a reasonable method of managing patients with asymptomatic tumors.
Surgery In his landmark 1957 paper, Simpson retrospectively graded the extent of meningioma resection in an attempt to find a correlation with recurrence (75). Table 3 details Simpson’s methodology in which surgical resection is graded from 1 (complete resection) to 5 (decompression only). Kinjo et al. further suggested a grade zero resection in which an additional 2 cm margin of dura is excised as a means for further reducing the rate of recurrence (76). From Simpson’s original work, the recurrence rate after surgical resection ranged from 9% for a grade I resection, 19% for grade II, 29% for grade III, to 44% for a grade IV resection (75). The extent of resection has since been strongly confirmed as the primary factor influencing meningioma recurrence rate. More recent studies have shown the 5-year progression-free survival to be between 77% and 93% after complete resection, and between 52% and 63% in patients with a subtotal resection. The 10year progression-free survival data ranges from 61% to 80% for a gross total resection and from 37% to 45% for a subtotal resection (56,77,78).
Adjuvant Therapy OUTCOME AND PROGNOSIS Asymptomatic Meningioma As the discovery of incidental meningiomas grows, understanding the natural history is becoming increasingly more important. One-third to two-fifths of meningiomas are asymptomatic (54,73), and several studies have assessed the growth rate of incidental meningiomas. Olivero et al. found that 10 of 45 patients with asymptomatic meningiomas exhibited tumor growth. Over an average imaging follow-up of 47 months, the average tumor growth in these 10 patients was 2.4 mm/yr (74). Yano et al. found that only 37% of asymptomatic meningiomas showed tumor growth and only 6% of patients became symptomatic over a mean follow-up of 3.9 years. Patients with tumors larger than 3 cm at diagnosis or T2-hyperintense tumors were more likely to become symptomatic over time while patients with calcified tumors were less likely to (73). In the subgroup of patients greater than 70 years old, the surgical morbidity associated with asymptomatic tumors was 9.4% as compared with 4.4% in patients less than 70 years. Further the surgical morbidity in this group exceeded the morbidity in the observation alone cohort (6%). Hence, for asymptomatic meningiomas, Yano et al. recommend serial neuroimaging and close clinical monitor-
Radiation therapy is an integral part of the treatment of meningiomas. In patients who have undergone a subtotal resection followed by adjuvant external beam radiation therapy, 5-year progression-free survival has been shown to be between 77% and 91% (56,59,79–81). In a retrospective analysis of 140 patients treated with subtotal resection followed by EBRT, Goldsmith et al. reported a 5-year progression-free survival rate of 85% for benign meningiomas and 58% for malignant tumors (59). Soyeur et al. more recently compared gross total resection, subtotal resection plus adjuvant EBRT, and subtotal resection followed by radiotherapy at tumor progression (56). Over a mean follow-up of 7.7 years, the 5year progression-free survival for gross total resection was 77%, while that for subtotal resection alone was 38%. Those patients who underwent subtotal resection and adjuvant radiotherapy had a 5-year progression-free survival rate of 91%. The overall survival for the three groups was not statistically different and was no different than the age-match general population. Current methods of treatment planning and delivery have led to decreased toxicity associated with EBRT than the 38% rate reported in earlier literature (82). The complication rate for radiation therapy is between 2.2% and 3.6% and includes cognitive decline, pituitary insufficiency, and radiation-induced neoplasms (59,82,83).
Chapter 36: Meningioma
Stereotactic radiosurgery (SRS) and stereotactic radiotherapy have shown promising results as both primary and adjuvant therapies for meningioma. Numerous retrospective studies since the 1990s have demonstrated 5-year local control rates with SRS of between 86% and 99%, tumor regression rates of 28% to 70%, and symptom improvement in 8% to 65% of patients (60). In their experience in treating patients with benign meningiomas less than 3.5 cm in average diameter, Pollock et al. showed radiosurgery to yield comparable results to those seen with a Simpson’s grade I surgical resection (84). However, compared with the population who underwent a Simpson’s grade II, III, or IV resection, stereotactic radiosurgery yielded a higher rate of progression-free survival (84). The 3 and 7 year rate of progression-free survival for stereotactic radiosurgery was 100% and 95% respectively, while that seen for Simpson’s grade I was 100% and 96%, Simpson’s grade II was 91% and 82%, and Simpson’s grade III and IV was 68% and 34% respectively. More recently, Kollova et al. reported their experience in treating 325 benign meningiomas with either primary or adjuvant SRS (85). Patients had a mean tumor volume of 4.4 cm3 and the authors achieved a tumor control rate of 97.9% at 5 years. Improvement in neurological symptoms such as imbalance, oculomotor palsy, trigeminal symptoms, hemiparesis, and vertigo occurred in 61.9% of patients. The permanent toxicity rate was 5.7%, which included seizures, trigeminal symptoms, hemiparesis, and others. Toxicity after radiosurgery is usually due to either symptomatic edema or cranial neuropathies. In particular, the special sensory nerves (optic and vestibulocochlear) appear the most sensitive (86). Vascular occlusion after stereotactic radiosurgery is a rare complication, but is estimated to occur in 1% to 2% of cases (87). The pathogenesis is thought to involve luminal narrowing after radiation-induced endothelial damage. As with stereotactic radiosurgery, stereotactic radiotherapy has shown high rates of progression-free survival of between 98% and 100% over a mean follow-up of 21 to 68 months (88,89). Studies have also shown an average reduction in tumor volume of 33% at 24 months and 36% at 36 months with stereotactic radiotherapy (90). Acute toxicities of stereotactic radiotherapy are generally mild and can include alopecia, skin erythema, and fatigue. The rate of late toxicity ranges between 2% and 13%. Late complications include hypopituitarism, visual deterioration, cognitive impairment, and tinnitus (60). Timing of adjuvant therapy for patients with benign meningioma who have undergone subtotal resection is still a matter of controversy. There is no randomized data to date to support postoperative radiation therapy versus radiation therapy at the time of tumor progression. The current Phase III prospective randomized control trial by the European Organisation of Research and Treatment for Cancer should aid in answering this question. REFERENCES 1. CBTRUS (2008). Statistical report: Primary brain tumors in the United States, 2000–2004. Central Brain Tumor Registry of the United States. Hinsdale, IL: 2008. 2. Hoffman S, Propp JM, McCarthy BJ. Temporal trends in incidence of primary brain tumors in the United States, 1985–1999. Neuro Oncol. 2006;8(1):27–37. 3. Vernooij MW, Ikram A, Tanghe HL. Incidental findings on brain MRI in the general population. N Eng J Med. 2007;357:1821– 1828. 4. Sloof J, Kernohan J, MacCarthy C. Primary Intramedullary Tumors of the Spinal Cord and Filum Terminale. Philadelphia, PA: WB Saunders, 1964.
509
5. Baumgartner JE, Sorenson JM. Meningioma in the pediatric population. J Neurooncol. 1996;29:223–228. 6. Rochat P, Johannesen HH, Gjerris F. Long-term follow-up of children with meningiomas in Denmark: 1935 to 1984. J Neurosurg. 2004;100:179–182. 7. Monteiro GT, Pereira RA, Kiofman RJ, et al. Head injury and brain tumors in adults: A case-control study in Rio de Janeiro, Brazil. Eur J Cancer. 2006;42(7):917–921. 8. Preston-Martin S, Paganini-Hill A, Henderson BE, et al. Casecontrol study of intracranial meningiomas in women in Los Angeles County, California. J Natl Cancer Inst. 1980;65(1):67– 73. 9. Preston-Martin S, Pogoda JM, Schlehofer B, et al. An international case-control study of adult glioma and meningioma: The role of head trauma. Int J Epidemiol. 1998;27:579–586. 10. Preston-Martin S, Yu MC, Henderson BE, et al. Risk factors for meningiomas in men in Los Angeles County. J Natl Cancer Inst. 1983;70(5):863–866. 11. Phillips LE, Koepsell TD, van Belle G, et al. History of head trauma and risk of intracranial meningioma: Population-based case-control study. Neurology. 2002;58:1849–1852. 12. Annegers JF, Laws ER Jr, Kurland LT, et al. Head trauma and subsequent brain tumors. Neurosurgery. 1979;4:203–206. 13. Inskip PD, Mellemkjaer L, Gridley G, et al. Incidence of intracranial tumors following hospitalization for head injuries (Denmark). Cancer Causes Control. 1998;9:109–116. 14. Al-Mefty O, Topsakal C, Pravdenkova S, et al. Radiationinduced meningiomas: Clinical, pathological, cytokinetic, and cytogenic characteristics. J Neurosurg. 2004;100:1002–1013. 15. Ron E, Modan B, Boice JD, et al. Tumors of the brain and nervous system after radiotherapy in childhood. N Engl J Med. 1988;319(16):1033–1039. 16. Shintani T, Hayakawa N, Hoshi M, et al. High incidence of meningioma among Hiroshima atomic bomb survivors. J Radiat Res. 1999;40:40–57. 17. Sadamori N, Shibata S, Mine M, et al. Incidence of intracranial meningiomas in Nagasaki atomic-bomb survivors. Int J Cancer. 1996;67(3):318–322. ¨ J, Kosteljanetz M, et al. Cellular tele18. Christensen HC, Schuz phones and risk for brain tumors: A population-based, incident case-control study. Neurology. 2005;64(7):1189–1195. ¨ J, Bohler ¨ 19. Schuz E, Berg G, et al. Cellular phones, cordless phones, and the risks of glioma and meningioma (Interphone Study Group, Germany). Am J Epidemiol. 2006;163(6):512–520. 20. Altieri A, Castro F, Bermejo JL, et al. Association between number of siblings and nervous system tumors suggests an infectious etiology. Neurology. 2006;67:1979–1983. 21. Topp WC, Lane D, Pollack R. Transformation by SV40 and polyomavirus. In: Tooze J, ed. DNA Tumor Viruses. New York, NY: Cold Spring Harbor Laboratory Press, Inc., 1981: 200–301. 22. Fanning E. Simian virus 40 large T antigen: The puzzle, the pieces, and the emerging picture. J Virol. 1992;66:1289–1293. 23. Zhen H, Zhang X, Bu X, et al. Expression of the simian virus 40 large tumor antigen (Tag) and formation of Tag-p53 and TagpRb complexes in human brain tumors. Cancer. 1999;86:2124– 2132. 24. Rollison DE, Utaipat U, Ryschkewitsch C, et al. Investigation of human brain tumors for the presence of polyomavirus genome sequences by two independent laboratories. Int J Cancer. 2005;113(5):769–774. 25. Simon M, Bostrom JP, Hartmann C. Molecular genetics of meningiomas: From basic research to potential clinical applications. Neurosurgery. 2007;60:787–798. 26. Wellenreuther R, Kraus JA, Lenartz D, et al. Analysis of the neurofibromatosis 2 gene reveals molecular variants of meningioma. Am J Pathol. 1995;146:827–832. 27. Hansson CM, Buckley PG, Grigelioniene G, et al. Comprehensive genetic and epigenetic analysis of sporadic meningioma for macro-mutations on 22q and micro-mutations within the NF2 locus. BMC Genomics. 2007;8:16. 28. Gutmann DH, Donohoe J, Perry A, et al. Loss of DAL-1 a protein 4.1-related tumor suppressor is an important early event in the pathogenesis of meningiomas. Hum Mol Genet. 2000;9:1495– 1500.
510
Viswanathan and DeMonte
29. Yi C, McCarty JH, Troutman SA, et al. Loss of the putative tumor suppressor band 4.1B/Dal1 gene is dispensible for normal development and does not predispose to cancer. Mol Cell Biol. 2005;25:10052–10059. 30. Riemenschneider M, Perry A, Reifenberger G. Histological classification and molecular genetics of meningiomas. Lacet Neurol. 2006;5:1045–1054. 31. Lamszus K. Meningioma pathology, genetics, and biology. J Neuropathol Exp Neurol. 2004;63(4):275–286. 32. Wrobel G, Roerig P, Kokocinski F, et al. Microarray-based gene expression profiling of benign, atypical and anaplastic meningiomas identifies novel genes associated with meningioma progression. Int J Cancer. 2005;114:249–256. 33. Wada K, Maruno M, Suzuki T, et al. Chromosomal and genetic abnormalities in benign and malignant meningiomas using DNA microarray. Neurol Reas. 2005;27:747–754. 34. Pecina-Slaus N, Martic TN, Tomas D. Meningiomas exhibit loss of heterozygosity of the APC gene. J Neuroonc. 2007, epub. 35. Smith JS, Lal A, Harman-Smith M, et al. Association between absence of epidermal growth factor receptor immunoreactivity and poor prognosis in patients with atypical meningioma. J Neurosurg. 2007;106;1034–1040. 36. Goldman CK, Bharara S, Palmer CA, et al. Brain edema in meningiomas is associated with increased vascular endothelial growth factor expression. Neurosurgery. 1997;40:1269–1277. 37. Provias J, Claffey K, delAguila L, et al. Role of vascular endothelial growth factor/vascular permeability factor in angiogenesis and peritumoral edema. Neurousurgery. 1997;40:1016– 1026. 38. Samoto K, Ikezaki K, Ono M, et al. Expression of vascular endothelial growth factor and its possible relation with neovascularization in human brain tumors. Cancer Res. 1995;55:1189– 1193. 39. Lamszus K, Lengler U, Schmidt NO, et al. Vascular endothelial growth factor, hepatocyte growth factor/scatter factor, basic fibroblast growth factor, and placenta growth factor in human meningiomas and their relation to angiogenesis and malignancy. Neurosurgery. 2000;46:938–947. 40. Bickerstaff ER, Small JM, Guest IA. The relapsing course of certain meningiomas in relation to pregnancy and menstruation. J Neurol Neurosurg Psychiatry. 1958;21:89–91. 41. Mehta D, Khatib R, Patel S. Carcinoma of the breast and meningioma. Association and management. Cancer. 1983;51:1937– 1940. 42. Hsu DW, Efird JT, Hedley-White ET. Progesterone and estrogen receptors in meningiomas: Prognostic considerations. J Neurosurg. 1994;86:113–120. 43. Rubinstein AB, Loven D, Geier A, et al. Hormone receptors in initially excised versus recurrent intracranial meningiomas. J Neurosurg. 1994;81:184–187. 44. Lusis EA, Chicoine MR, Perry A. High throughput screening of meningioma biomarkers using a tissue microarray. J Neurooncol. 2005;73:219–223. 45. Pravdenkova S, Al-Mefty O, Sawyer J, et al. Progesterone and estrogen receptors: Opposing prognostic indicators in meningiomas. J Neurosurg. 2006;105:163–173. 46. Perry A, Louis DN, Scheithauer BW, et al. Meningiomas. In: Kleihues P, Cavanee WK, eds. Pathology and Genetics of Tumors of the Nervous System: World Health Organization Classification of Tumors Lyon, France: IARC Press, 2007. 47. Perry A, Scheithauer BW, Stafford SI, et al. “Rhabdoid” meningioma: An aggressive variant. Am J Surg Pathol. 1998;22;1482– 1490. 48. Perry A, Scheithauer BW, Stafford SI, et al. “Malignancy” in meningiomas: A clinicopathological study of 116 patients, with grading implications. Cancer. 1999;85:2046–2056. 49. Schnitt SJ, Vogel H. Meningiomas. Diagnostic value of immunoperoxidase staining for epithelial membrane antigen. Am J Surg Pathol. 1986;10:640–649. 50. Artlich A, Schmidt D. Immunohistochemical profile of meningiomas and their histological subtypes. Hum Pathol. 1990;21:843–849.
51. Bruna J, Brell M, Ferrer I, et al. Ki-67 proliferative index predicts clinical outcome in patients with atypical or anaplastic meningioma. Neuropathology. 2007;27:114–120. 52. Matsuno A, Hiroshi N, Nagashima T. Histopathological analysis and proliferative potentials of intracranial meningiomas using bromodeoxyuridine and MIB-1 immunohistochemistry. Acta Histochem Cytochem. 2005;38:9–15. 53. Ribalta T, McCutcheon I, Aldape K, et al. The mitosis-specific antibody anti-phosphohistone-H3 (PHH3) facilitates rapid reliable grading of meningiomas according to WHO 2000 criteria. Am J Surg Pathol. 2004;28:1532–1536. 54. Drummond KJ, Zhu J, Black PM. Meningiomas: Updating basic science, management and outcome. Neurol. 2004;10(3):113– 130. 55. Cushing H, Eisenhardt L. Meningiomas: Their Classification, Regional Behaviour, Life History, and Surgical End Results. Springfield, IL: Charles C Thomas, 1938. 56. Soyuer S, Chang EL, Selek U, et al. Radiotherapy after surgery for benign cerebral meningioma. Radiother Oncol. 2004;71:85– 90. 57. Turbin RE, Thompson CR, Kennerdell JS. A long-term visual outcome comparison in patients with optic nerve sheath meningioma managed with observation, surgery, radiotherapy, or surgery and radiotherapy. Ophthalmology. 2002;109:890–900. 58. Rogers L, Mehta M. Role of radiation therapy in treating intracranial meningiomas. Neurosurg Focus. 2007;23(4):E4. 59. Goldsmith BJ, Wara WM, Wilson CB. Postoperative irradiation for subtotally resected meningiomas. A retrospective analysis of 140 patients treated from 1967 to 1990. J Neurosurg. 1994;80:195– 201. 60. Elia AE, Shih HA, Loeffler JS. Stereotactic radiation treatment for benign meningiomas. Neurosurg Focus. 2007;23(4):E5. 61. Khoo VS, Adams EJ, Saran F, et al. A comparison of clinical target volumes determined by CT and MRI for the radiotherapy planning of base of skull meningiomas. Int J Radiat Oncol Biol Phys. 2000; 46(5):1309–1317. 62. Schrell UM, Rittig MG, Anders M, et al. Hydroxyurea for treatment of unresectable and recurrent meningiomas. I. Inhibition of primary human meningioma cells in culture and in meningioma transplants by induction of the apoptotic pathway. J Neurosurg. 1997;86:845–852. 63. Loven D, Hardoff R, Sever ZB, et al. Non-resectable slowgrowing meningiomas treated by hydroxyurea. J Neurooncol. 2004;67:221–226. 64. Newton H, Slivka M, Stevens C. Hydroxyurea chemotherapy for unresectable or residual meningioma. J Neurooncol. 2000;49:165–170. 65. Ragel BT, Gillespie DL, Kushnir V, et al. Calcium channel antagonists augment hydroxyurea- and RU486-induced inhibition of meningioma growth in vivo and in vitro. Neurosurgery. 2006;59:1109–1121. 66. Zhang Z, Wang J, Muhr C, et al. Synergistic inhibitory effects of interferon-alpha and 5-fluorouracil in meningioma cells in vitro. Cancer Lett. 1996;100:99–105. 67. Kaba SE, DeMonte F, Bruner JM, et al. The treatment of recurrent unresectable and malignant meningiomas with interferon alpha-2B. Neurosurgery. 1997;40:271–275. 68. Chamberlain MC, Tsao-Wei DD, Groshen S. Temozolomide for treatment-resistant recurrent meningioma. Neurology. 2004;62:1210–1212. 69. Matsuda Y, Kawamoto K, Kiya K, et al. Antitumor effects of antiprogesterones on human meningioma cells in vitro and in vivo. J Neurosurg. 1994;80:527–534. 70. Grunberg S, Weiss M, Spitz I, et al. Treatment of unresectable meningiomas with the antiprogesterone agent mifepristone. J Neurosurg. 1991;74:861–866. 71. Gupta V, Su YS, Samuelson CG, et al. Irenotecan: A potential new chemotherapeutic agent for atypical or malignant meningiomas. J Neurosurg. 2007;106:455–462. 72. Chamberlain MC, Tsao-Wei DD, Groshen S. Salvage chemotherapy with CPT-11 for recurrent meningioma. J Neurooncol. 2006;78:271–276.
Chapter 36: Meningioma 73. Yano S, Kuratsu J. Kumamoto Brain Tumor Research Group. Indications for surgery in patients with asymptomatic meningiomas based on an extensive experience. J Neurosurg. 2006;105:538–543. 74. Olivero WC, Lister JR, Elwood PW. The natural history and growth rate of asymptomatic meningiomas: A review of 60 patients. J Neurosurg. 1995;83(2):222–224. 75. Simpson D. The recurrence of intracranial meningiomas after surgical treatment. J Neurol Neurosurg Psychiatry. 1957;20:22– 39. 76. Kinjo T, Al-Mefty O, Kanaan I. Grade zero removal of supratentorial convexity meningiomas. Neurosurgery. 1993;33(3):394– 399. 77. Condra KS, Buatti JM, Mendenhall WM, et al. Benign meningiomas: Primary treatment selection affects survival. Int J Radiat Oncol Biol Phys. 1997;39:427–436. 78. Mirimanoff RO, Dosoretz DE, Linggood RM, et al. Meningioma: Analysis of recurrence and progression following neurosurgical resection. J Neurosurg. 1985;80:191–194. 79. Barbaro NM, Gutin PH, Wilson CB, et al. Radiation therapy in the treatment of partially resected meningiomas. Neurosurgery. 1987;20:525–528. 80. Miralbell R, Linggood RM, De la Monte S, et al. The role of radiotherapy in the treatment of subtotally resected benign meningiomas. J Neurooncol. 1992;13:157–164. 81. Taylor BW Jr, Marcus RB Jr, Friedman WA, et al. The meningioma controversy: Postoperative radiation therapy. Int J Radiat Oncol Biol Phys. 1988;15:299–304. 82. Al-Mefty O, Kersh JE, Routh A, et al. The long-term side effects of radiation therapy for benign brain tumors in adults. J Neurosurg. 1990;73:502–512. 83. Glaholm J, Bloom HJG, Crow JH. The role of radiotherapy in the management of intracranial meningiomas: The Royal Marsden Hospital experience with 186 patients. Int J Radiat Oncol Biol Phys. 1990;18:755–761.
511
84. Pollock BE, Stafford SL, Utter A, et al. Stereotactic radiosurgery provides equivalent tumor control to Simpson Grade 1 resection for patients with small- to medium-size meningiomas. Int J Radiat Oncol Biol Phys. 2003;55:1000–1005. 85. Kollova A, Liscak R, Novotny J Jr, et al. Gamma knife surgery for benign meningioma. J Neurosurg. 2007;107(2):325– 336. 86. Tishler RB, Loeffler JS, Lunsford LD, et al. Tolerance of cranial nerves of the cavernous sinus to radiosurgery. Int J Radiat Oncol Biol Phys. 1993;27:215–221. 87. Barami K, Grow A, Brem S, et al. Vascular complications after radiosurgery for meningiomas. Neurosurg Focus. 2007;22(3):E9. 88. Debus J, Wuendrich M, Pirzkall A, et al. High efficacy of fractionated stereotactic radiotherapy of large base-of-skull meningiomas: Long-term results. J Clin Oncol. 2001;19:3547– 3557. 89. Jalali R, Loughrey C, Baumert B, et al. High precision focused irradiation in the form of fractionated stereotactic radiotherapy (SCRT) for benign meningiomas predominantly in the skull base location. Clin Oncol (R Coll Radiol). 2002;14:103– 109. 90. Henzel M, Gross M, Hamm K, et al. Significant tumor volume reduction of meningiomas after stereotactic radiotherapy: Results of a prospective multicenter study. Neurosurgery. 2006;59:1188– 1194. 91. Gautier-Smith C, Lamouche M. Parasagittal and Falx Meningiomas. London: Butterworths, 1987. 92. Quest D. Meningiomas: An update. Neurosurgery. 1978;3:219– 225. 93. MacCarty C, Taylor W. Intracranial meningiomas: Experiences at the Mayo Clinic. Neurologica Medico-Chir (Tokyo). 1979;19:569–574. 94. Rohringer M, Sutherland GR, Louw DF, et al. Incidence and clinicopathological features of meningioma. J Neurosurg. 1989;71:665–672.
37 Schwannomas of the Skull Base Daniel W. Nuss and Emily Lifsey Burke
considered in aggregate). Rarely, schwannomas involve the hypoglossal nerve, and extraocular motor nerves. Other sites have been reported as well; these are discussed in more depth later in the chapter. Both the genders are equally affected across the gamut of most anatomic sites, but intracranial schwannomas affect women more often than men, in a ratio estimated at 3:2 to 2:1. Schwannomas can occur at any age, but 75% occur in the third to fourth decades of life. There is no racial predisposition for schwannomas (3,5). Epidemiologically, it is essential to recognize that schwannomas occur across a spectrum of clinical scenarios. Certainly, schwannomas occur sporadically as isolated tumors in individuals with no history of familial disease, but they are more frequent in the setting of the neurofibromatosis syndromes. Therefore, a patient with a newly diagnosed or suspected schwannoma must be carefully evaluated for the possible presence of one of the genetic neurofibromatosis syndromes. This is true not only because early diagnosis is desirable, but because prognosis depends on accurate assessment, and the treatment options and decision-making are more complicated when multiple tumors are present, or expected to develop over the patient’s lifetime.
INTRODUCTION The schwannoma, in the simplest terms, is a tumor of nervesheath cells. While schwannomas of the skull base are uncommon, they are not rare. Intuitively this makes sense, given the rich network of nerves in the head and neck region. In fact, about 25% to 45% of all reported schwannomas are found in the head and neck (1,2), and since most of these originate from cranial nerves, a significant proportion will involve the skull base. Therefore, most practitioners in otolaryngology, neurosurgery, and related specialties who care for patients with skull base problems will encounter schwannomas in clinical practice. Despite their generally benign nature, schwannomas at or near the skull base can be responsible for life-changing morbidity due to associated neurologic deficits, and also due to frequent association with the numerous manifestations of the neurofibromatosis syndromes. They can be lethal as well, because of proximity to, or encroachment upon, neurovascular structures and the airway. Successful management of these benign neoplasms demands a sophisticated understanding of their origin, behavior, clinical features, and treatment options, along with clinical judgment that takes into consideration the natural history of the lesion as well as the likely morbidity of therapy. The purpose of this chapter is to provide the reader with a practical overview of the incidence, pathology, clinical features, treatment options, and outcomes for management of skull base schwannomas. Vestibular (eighth cranial nerve) schwannomas, discussed elsewhere in this text, will be excluded from this discussion. Specific surgical techniques, also not the focus of this chapter, will be addressed elsewhere.
The Neurofibromatoses There are now at least three major types (and eight or more subtypes) of neurofibromatosis syndromes: type 1 (NF1), type 2 (NF2), and the recently described entity referred to as schwannomatosis (6–9). As a matter of perspective, the NF1 syndrome is much more common than NF2. Worldwide, the incidence of NF1 is approximately 1 in 4000, irrespective of race, gender, or ethnic background. The incidence of NF2 has recently been estimated at around 1 in 25,000 (10); this recent estimate is higher than historical reports, likely because of increasing awareness of diagnostic criteria and perhaps in part because of improved diagnostic imaging. Because schwannomatosis has only recently been recognized as a separate entity, its real incidence is not yet well understood. A Finnish population-based review in 2000 (11) estimated the annual incidence at 1 in 1,700,000, but MacCollin et al. (12) more recently postulated that it may be just as common as NF2. All of these syndromes reflect an underlying genetic abnormality, discussed in detail below, which predisposes patients to develop tumors of the nervous system (see “Genetics and Molecular Features”). Specific criteria for diagnosis of NF1 and NF2, promulgated by The National Institutes of Health (13), are as follows:
INCIDENCE AND EPIDEMIOLOGY While the exact incidence and prevalence of benign schwannomas in all sites are not known, it is clear that nerve sheath tumors make up 8% to 10% of all intracranial, extra-axial tumors (3). Of those, schwannomas are the majority (5). As noted above, 25% to 45% of all schwannomas are found in the head and neck region. Although this discussion will focus on schwannomas affecting the skull base, it should be noted that schwannomas can occur in many other sites in the head and neck as well (face, skin, orbit, lips, maxilla, mandible, oral cavity, sinuses, parotid, nasopharynx, and larynx) (5). Schwannomas in general have a predilection for sensory nerves, although they do also occur along motor nerves. They arise most commonly (in descending order) from the vestibular component of the eighth nerve (>90%), sensory divisions of the trigeminal nerve, the facial nerve, and the nerves of the jugular foramen (cranial nerves IX, X, and XI
The criteria for the diagnosis of NF1 are met in an individual if two or more of the following signs are found: 513
514
Nuss and Burke
r Six or more caf´e au lait macules larger than 5 mm in the greatest diameter in prepubertal children and larger than 1.5 cm in postpubertal individuals r Two or more neurofibromas of any type, or 1 plexiform neurofibroma r Multiple freckles (Crowe sign) in the axillary or inguinal region r A distinctive osseous lesion, such as sphenoid dysplasia or thinning of long bone cortex, with or without pseudoarthrosis r Optic glioma r Two or more iris hamartomas (Lisch nodules) seen on slit-lamp or biomicroscopy examination r A first-degree relative (parent, sibling, offspring) with NF1, as diagnosed by using the above criteria. To diagnose NF2, the following criteria must be met: r Bilateral vestibular schwannomas, or r A family history of NF2 (parent, sibling, or child) plus a unilateral vestibular schwannoma before age 30 r or any two of the following: ◦ glioma ◦ meningioma ◦ schwannoma ◦ juvenile posterior subcapsular lenticular opacity (juvenile cortical cataract).
Schwannomatosis In recent years, the separate clinical entity known as schwannomatosis has been defined (6–8,12,14–16). According to MacCollin and colleagues, “schwannomatosis is a recently recognized third major form of neurofibromatosis that causes multiple schwannomas without vestibular tumors.” Diagnosis is based on the presence of multiple schwannomas without the stigmata of neurofibromatosis NF1 or NF2. It has been estimated that patients with schwannomatosis may represent up to 5% of all patients requiring schwannoma resection (12). Epidemiologically the condition may be just as common as NF2, but for unclear reasons, it is not apparently familial. Schwannomatosis patients have been noted to have tumors of numerous sites, including intraspinal, paraspinal, brachial plexus, femoral nerve, sciatic nerve, calf, forearm, retroperitoneum, and middle cranial/infratemporal fossa region. The common presenting symptoms included paresthesias, palpable mass, pain, or weakness, with pain being the dominant clinical problem and indication for surgery. Huang et al., in reporting a series of six schwannomatosis patients, concluded that surgery is indicated for symptomatic lesions, while asymptomatic tumors should be followed conservatively (17). Because of the increased risk for developing multiple schwannomas, they recommend regular surveillance and consideration of genetic counseling, even though the exact mechanism of genetic alteration is unknown. The relative paucity of information on this condition makes it especially important that such patients be studied prospectively.
Associated Conditions and Considerations In discussing schwannomas, neurofibromas, and their associated NF syndromes, it should be emphasized that persons who have any of these clinical syndromes are members of populations who are at risk for other very serious problems, including tumors and nontumor conditions as well. Other tumors that occur with more frequency in these syndromes include gliomas of the optic nerve, astrocytoma, meningioma, intramedullary glioma, ependymoma, soft tissue sarcomas,
and juvenile myeloid leukemia. Also, some individuals with neurofibromatosis have below average intelligence; 25% to 40% have learning disabilities (e.g., attention-deficit hyperactivity disorder/ADHD, neuromotor dysfunction, visualspatial processing disorders); and 5% to 10% have mental retardation. Certain endocrine problems can also be associated with NF. Short stature, growth hormone deficiency, sexual precocity, and pheochromocytomas all can occur more commonly in NF patients than in the general population (1,18). Care of patients with any of the neurofibromatoses, for all the above reasons, crosses multiple disciplines and requires a high level of awareness among all involved clinicians.
PATHOLOGY Definition and Nomenclature Historically, many different names have been used to describe what is now classified as schwannoma, including neuroma, neurinoma, neurilemmoma, neurolemmoma, perineural fibroblastoma, and others. These older terms are now considered inaccurate and their use is discouraged, but they are acknowledged here since they do appear in older but stillimportant reference texts and clinical literature.
Distinguishing Between Schwannomas and Neurofibromas In any discussion of schwannomas, it is essential to distinguish the schwannoma from the one other common nerveassociated tumor that affects the skull base, namely the neurofibroma. They are both encountered with some regularity in clinical practice, yet the prognostic implications as well as the therapeutic decision-making are sometimes significantly different. Although schwannomas and neurofibromas do share certain clinical characteristics, as well as a common cell-type in the progenitor Schwann cell, the two entities are really quite distinct. Clinicians and pathologists alike must be mindful of the features that set them apart, for those features straddle both clinical and pathological lines. In actual practice, when all of the clinical and pathological features are taken into account, precise diagnosis is usually straightforward. Clinically, schwannomas are usually slow-growing masses that may be sensitive or painful, especially when they are in locations subject to pressure or manipulation (e.g., the upper neck or parapharyngeal space), and they are often tender to palpation. Neurofibromas are usually asymptomatic. The Schwann cell is the parent cell of both schwannomas and neurofibromas. The Schwann cell is a sheath cell, not a nerve cell. Schwann cells, which can be regarded as the nervous system’s cellular maintenance team, are important for nerve sheath integrity and myelin production. According to Batsakis (1), the Schwann cell is regarded by many neuropathologists as homologous with the oligodendroglia of the central nervous system, and it is believed to be derived from the neural crest, and therefore considered as neuroectodermal in origin. This is supported by immunohistochemical findings (see below), with schwannoma strongly positive for S-100 protein, which is helpful in distinguishing schwannoma from other soft-tissue tumors. Pathologically, the typical schwannoma is an encapsulated, solitary, and expansile tumor that is often fusiform and attached to a recognizable nerve (Fig. 1). On cut section, they are usually mostly white or yellowish-white and firm. The tumor appears to push axons aside rather than enveloping them (“centrifugal” growth), at least early on in its growth. Initially fusiform, these tumors at the skull base tend to assume the shape of the confining space in the area of origin,
Chapter 37: Schwannomas of the Skull Base
Figure 1 Gross specimen of resected schwannoma from parapharyngeal space at base of skull. Note fusiform shape tapering to nerve trunk on each end.
and then progressively erode adjacent bone. When in proximity to a foramen or fissure of the skull, they assume a bi-lobed or “dumbbell” shape with expansion on either side of the bony threshold [Fig. 2(A)and 2(B)]. Usually there is a distinctive biphasic histological pattern of well-developed cylindrical structures (known as Antoni type A tissue), with palisading nuclei around a central mass of cytoplasm, the so-called Verocay body (1,5,19). These well-recognized features often are embedded within a more loosely textured, less well-developed, hypocellular stroma in which fibers and cells form no distinctive pattern (Antoni B tissue) (1). Another very common feature of schwannomas is their rich vascularity. As discussed below, this gives rise to certain distinctive histological and radiographic characteristics, as well as the tendency for schwannomas to bleed. Fine pericellular reticulin can be seen with special stains. Histopathological findings of schwannoma are depicted in Figure 3. “Retrogressive features” (1) are very common in schwannomas, including necrosis, areas of cystic degeneration, and vascular changes. Owing to their rich supply of blood vessels that have a tendency to be abnormal, schwannomas usually feature angiomatous clusters of blood vessels that are prone to focal thrombosis, and areas of recent and old hemorrhage. The term “cystic schwannoma” is applied to tumors that, because of the retrogressive/degenerative
515
phenomena, are filled with serous fluid. The related term, “ancient schwannoma” describes tumors in which there is extensive hyalinization, again, believed to be related to degenerative phenomena. These kinds of features can often be appreciated even on examination of the bisected gross tumor, but are more apparent at light microscopy. Sometimes these retrogressive features, because of the resulting inhomogeneity within a given tumor, can be misinterpreted to represent malignancy, which is extremely rare in true schwannomas (see discussion on malignancy below), and must be evaluated with caution. Neurofibromas do not share most of the above features. Neurofibromas do not have the collagenous capsule that schwannomas have; they are instead surrounded by thickened perineurium and epineurium. They incorporate axons into the tumor and envelop them (“centripetal” growth). Typically they are more cellular, manifesting a spindle cell pattern of growth. They do not often exhibit retrogressive features. Neurofibromas do not generally display the Antoni type A and B patterns, nor do they have Verocay bodies as schwannomas do. Instead, the matrix of the tumor is made up of type IV collagen fibers and myxoid material. The interspersed cell populations (Schwann cells, fibroblasts, and perineural cells) are scattered within this matrix. Many are seen in association with the overall clinical picture of generalized neurofibromatosis. And as discussed below, malignant degeneration is well established with neurofibromas.
Immunohistochemistry and Special Stains As derivatives of embryonic neural crest cells, Schwann cells are of neuroectodermal origin. Consequently, tumors derived from these cells will stain strongly positive for S-100 protein on immunohistochemistry. While S-100 protein is not specific for schwannoma, it is helpful in diagnosis, since schwannomas will stain much more heavily for S-100 than will neurofibromas. (Other cell types that are S-100 positive include glial cells, melanocytes, chondrocytes, adipocytes, myoepithelial cells, macrophages, Langerhans cells, dendritic cells, and keratinocytes, all of which are of neuroectodermal origin.) The immunoreactivity for S-100 protein is seen in fewer cells making up the neurofibroma, as opposed to more blanketed reactivity of the schwannoma. Trichrome and Alcian blue stains will highlight the collagen and mucinous matrix of neurofibromas (3,5,18,19). Schwannomas may also variably stain positive for GFAP (glial fibrillary acid protein), keratin, and EMA (epithelial membrane antigen), but because of this variability, they are rarely helpful in diagnosis.
Genetics and Molecular Features
Figure 2 (A) Parasagittal T1-weighted MR image of trigeminal schwannoma demonstrating bilobed or “dumbbell ” shape typical of schwannoma that develops on both sides of a foramen or fissure. Note that the tumor is heterogeneous in its signal characteristics, another common feature of schwannomas (see text). (B) Axial T2-weighted MR image of same patient, showing dumbbell shape with smaller portion of tumor in posterior fossa (slightly compressing brainstem), and larger portion in middle fossa. Tumor demonstrates higher signal intensity on T2 series than on T1 (see text).
Much has been learned in recent years about the genetics of neurofibromatosis, and in aggregate, discoveries concerning the neurofibromatosis syndromes have provided unique insight into the genetic basis of tumor formation in general. Neurofibromatosis type 1 (NF1) has been traced to a defect in the neurofibromin gene and has an autosomal dominant inheritance pattern. This is a large gene located on chromosome 17. The fact that the gene is of large size may be the cause of the high number (40%) of sporadic cases of NF1; defects in the germ cell line cause the sporadic cases of disease. The neurofibromin protein normally functions to inhibit Ras, which is commonly activated in human cancers. The loss of function of neurofibromin causes uninhibited activity of Ras and thus unregulated cell growth. NF1, also known as von Recklinghausen disease, results in neurofibromas of cranial and (more commonly) peripheral nerves, frequently arising
516
Nuss and Burke
Figure 3 Histopathological sections demonstrating typical features of schwannoma. Biphasic pattern (A) with Verocay bodies often seen in the compact Antoni A type tissue, along with pallisading nuclei (B). (C) Nuclear pleomorphism, which can be seen as part of so-called ancient change, sometimes mistaken as features of malignancy. (D) Microcystic changes, inflammation, and hemosiderin characteristic of schwannoma. (E) Strong, diffusely positive stain for S-100 protein. (F) Fine pericellular reticulin. Source: From Ref. 19.
in the skin. Overall, the penetrance of NF1 is variable, from very mild to severe. Neurofibromatosis type 2 is caused by a defect in the merlin or schwannomin gene, located on chromosome 22. The loss of chromosome 22q is seen in sporadic cases of schwannomas, but the NF2 defect usually causes premature truncation or complete deletion of the entire gene. Schwannomin acts differently than neurofibromin, in that it is a membrane/cytoskeleton protein involved in cell motility and proliferation, but it also affects Ras/Rac activity. The hallmark of this disorder is bilateral vestibular schwannomas, and affected persons may also have peripheral schwannomas and meningiomas. There is a spectrum of disease, but in general, NF2 is considered a more severe and morbid disease than NF1. The newer entity of schwannomatosis has also been traced to a defect in the schwannomin gene on chromosome
22 as its cause. What distinguishes this from NF2 has not yet been determined, but the two seem to have significant molecular overlap. In all of the neurofibromatosis syndromes, the affected genes basically fail to perform in their protective role as tumor suppressors. NF-related tumorigenesis takes place in what has become known as the “two-hit” model of genetic disease. As such, one copy of the gene, inherited from a parent, carries the mutation, but when another “hit,” or spontaneous mutation on the normal (second) copy occurs, the end result is a failure of tumor suppression. Eventually tumors become clinically manifest. While a more detailed discussion of the molecular genetics of neurofibromatosis is beyond the scope of this chapter, the molecular basis of these disease processes is fascinating and has provided invaluable information on tumor biology (14,18,19).
Chapter 37: Schwannomas of the Skull Base
517
Figure 4 Malignant peripheral nerve sheath tumor. (A) Clinical photograph of patient with NF type I, who experienced sudden and dramatic growth of longstanding facial tumor, ultimately diagnosed as malignant peripheral nerve sheath tumor. (B) CT of the same patient showing deeply infiltrating nature of tumor with invasion of temporal fossa, orbit, skull, temporomandibular joint and facial soft tissues.
Malignancy and Schwannomas Although all neurofibromatosis patients will develop benign tumors, approximately 5% to 10% will also develop malignant nerve-sheath tumors. A smaller number will develop other malignancies including leukemia and soft tissue sarcomas. Therefore, appropriate suspicion and surveillance for malignancy must be part of their regular follow-up. The nomenclature of nerve-sheath malignancies is sometimes a source of confusion and even among experts is not without some controversy. Pathologists and surgeons have historically applied the terms differently, leading to many clinical descriptions of so-called “malignant schwannoma” in cases for which pathologists would prefer to use other terms. The explanation for this relates to the process by which these malignant tumors develop. Undoubtedly the majority of malignant neoplasms of the peripheral nerves are of Schwann cell origin. However, the majority of the pathological literature on the subject dismisses any relationship between the benign schwannoma and any malignant degeneration. In other words, benign schwannomas do not degenerate into malignant tumors, or at least if they do it happens exceedingly rarely (1,5,20), and it is nearly always in patients who have neurofibromatosis (21). Therefore, the term “malignant schwannoma,” technically speaking, should rarely be used, even though it often appears in clinical case reports. “Neurogenous sarcoma” and more recently “malignant peripheral nerve sheath tumor” (MPNST) have been proposed as more accurate terms for malignant tumors of peripheral nerves (1,5). While schwannomas are rarely if ever associated with malignant degeneration, neurofibromas frequently are. Malignant nerve-sheath tumors have clearly been linked to neurofibromatosis, and have consistently been observed evolving from established neurofibromas, especially in the clinical setting of generalized neurofibromatosis (as opposed to solitary neurofibroma cases). The incidence of malignancy in association with neurofibromas is generally accepted to be around 8%, with estimates ranging from 5% to 16.4% (22,23). Malignant peripheral nerve-sheath tumors (MPNSTs) are most appropriately classified as highly aggressive softtissue sarcomas, and are often lethal. Their clinical course is different from most schwannomas and neurofibromas. Unlike the benign nerve-sheath lesions, malignant ones often
present with rapidly growing, painful tumors. They may reach extreme proportions in short periods of accelerated growth [Fig. 4(A)and 4(B)]. Because this chapter is devoted to benign skull base schwannomas, a more detailed discussion of malignant peripheral nerve tumors will not be included here. The reader is referred to several excellent references for more information on that subject (24–26).
RADIOGRAPHIC FEATURES OF SKULL BASE SCHWANNOMAS As with all skull base lesions, radiographic imaging of schwannomas must be thorough in order to give the best chance of accurate diagnosis. Both CT and MRI should be obtained, as their findings are complimentary; images should be obtained in three planes.
Features Common to Both CT and MR On both CT and MRI, schwannomas may be heterogeneous because of their tendency for cystic degeneration and intratumoral hemorrhage, as detailed above in the section on pathology. Smaller tumors (less than 1.5 cm) tend to be mostly solid, but larger ones will have both cystic and solid areas (27). The solid component of a schwannoma will enhance brightly with contrast due to inherent vascularity, but the cystic components may not. Most schwannomas will be well-defined lesions with smooth borders. The dumbbell shape, along with smooth expansion of bony foramina, is highly suggestive of, if not pathognomonic for, schwannoma [Figs. 2(A), 2(B), 5, and 6].
CT Characteristics CT images typically demonstrate lesions that are isodense to hyperdense compared to muscle, although cystic components will be hypodense. Thin sections are essential to accurately assess bone erosion and related bony anatomy of the adjacent skull. Most schwannomas will erode bone gradually and smoothly, enlarging foramina of affected nerves, and causing remodeling of adjacent bone as tumor expands. Often there will be a sclerotic rim of bone at the margins. In many cases, the affected foramen will be greatly expanded
518
Nuss and Burke
MR Characteristics On MRI, most schwannomas will appear hypo- to isointense relative to brain (white matter) on T1 sequences, and iso- to hyperintense on T2 sequences (27,28) [Figs. 2(A), 2(B), and 5(C)].
CLINICAL FEATURES OF SKULL BASE SCHWANNOMAS General Considerations Regarding Nonvestibular Schwannomas Anatomic Distribution Nonvestibular schwannomas (NVSs) (i.e., tumors of nerves other than cranial nerve VIII) have been reported to involve virtually every nerve associated with the skull base, including all of the cranial nerves as well as the sympathetic, parasympathetic, and upper cervical nerves. They may develop as purely intracranial, purely extracranial, or transcranial entities. The clinical symptoms and differential diagnosis, discussed in the sections that follow, vary with the specific site of origin and direction of growth.
Natural History
Figure 5 (A) Axial CT and (B) axial T1-weighted MR of patient with V2 schwannoma. In this case, the tumor has widened the pterygomaxillary fissure, with bilobed growth into infratemporal fossa and nasal cavity. (C) Coronal T2-weighted MR demonstrating dumbbell tumor of V2 (maxillary division of trigeminal) straddling pterygomaxillary fissure on patient’s right side. Arrow points to the normal V2 on the patient’s left side. V2 is not visible on right side because it is the nerve of origin of tumor.
such that it may be difficult to recognize, or may appear to be absent (Figs. 5 and 6). Ancient schwannomas may show irregular calcification.
Patients with neurofibromatosis type 2 (NF2) are at particularly high risk of developing schwannomas, and much can be learned about schwannomas by examining this population. As part of the NF2 Natural History Consortium project, Fischer et al. (29) studied the prevalence and location of nonvestibular cranial nerve schwannomas among NF2 patients. Magnetic resonance imaging findings were prospectively gathered for 83 patients, over three consecutive annual evaluations. More than half (51%) were found to have nonvestibular schwannomas (NVSs) of the cranial nerves. Of these, 25 (60%) also had cranial meningiomas, and 21 of those without NVS (25% of 83) had at least 1 meningioma. Several other interesting observations came from Fischer’s review. The average size of the incidental NVS was small, at 0.4 cm3 , and overall, there was no significant change in NVS during the roughly 3 year interval of the study. The most common locations of the NVSs in this NF2 population were oculomotor and trigeminal. Neuropathies associated with tumors of the upper cranial nerves were few in this series. In contrast, lower cranial nerve schwannomas were
Figure 6 CT-derived images of a 4-year-old patient with V3 tumor, showing extensive smooth expansion of foramen ovale, typical for schwannoma. (A) and (B), Three-dimensional CT reconstructions showing intracranial and extracranial views, respectively. Arrows point to normal foramen ovale on unaffected side. (C) Coronal CT scan showing extensive expansion of foramen ovale and also considerable remodeling of adjacent mandible due to tumor in infratemporal fossa.
Chapter 37: Schwannomas of the Skull Base
often symptomatic, associated with swallowing difficulty, aspiration, and other sequelae. Mautner, Samii, and colleagues studied the clinical spectrum of disease, and radiographic findings, in a population of 48 NF2 patients, using a protocol that included gadolinium-enhanced MRI of the brain and spine as well as neurological, dermatological, and ocular examinations (30). Their study revealed that 96% had vestibular schwannomas, 90% had spinal tumors, 63% had subcapsular cataracts, 58% had meningiomas, and 29% had trigeminal schwannomas. They emphasized that the incidence of spinal tumors, a major source of morbidity and mortality, was higher in this report than in previous studies. These reviews illustrate that NF2 patients are highly likely to develop nonvestibular schwannomas, and that these patients are also very likely to develop multiple nonschwannoma tumors (especially spinal tumors, neurofibromas, and meningiomas) over time. This latter observation underscores the importance of long-term clinical follow-up with serial imaging, and it also emphasizes the need for careful consideration of all treatment options, including observation, for newly diagnosed schwannomas in NF2 patients.
Schwannomas in Children For pediatric patients with schwannomas in the setting of NF2, the multiplicity and variety of tumors—and the resulting morbidity—appears to be even worse than for adults. Nunes and MacCollin reported their findings from a series of 12 patients with neurofibromatosis 2 (31). One-third of the patients presented with hearing loss and another third presented with other cranial neuropathies. Tumor-related disability in many patients was high, with documented cranial meningiomas in 75%, cranial schwannomas other than vestibular schwannomas in 83%, and spinal cord tumors in 75%. Functionally, 75% of children had hearing loss, 83% had visual impairment, 25% had abnormal ambulation, and 25% were performing below their academic grade level. The authors concluded that our ability to diagnose pediatric patients with NF2 is improving, but outcomes appear to be significantly worse than in adult patients, and that work is needed to determine optimal management of pediatric neurofibromatosis 2.
Symptoms Vary with Location Schwannomas of the skull base exhibit a wide variation in the degree to which they produce signs and symptoms. While some tumors are essentially asymptomatic, others produce local symptoms that are directly attributable to the tumor’s nerve of origin. Still others become symptomatic by virtue of the pressure they place on surrounding structures, while the nerve of origin may remain free of any demonstrable signs of deficit until later in the course of progression. One of the chief determinants of a tumor’s clinical impact is whether the tumor volume is mostly intracranial, mostly extracranial, or a combination of both. In fact, symptoms may be more related to the tumor’s pattern of growth than to its nerve of origin. As a general rule, when the tumor’s volume is mostly intracranial, the patient is more likely to have symptoms that are global or due to central nervous system effects; when most of the tumor’s volume is extracranial, the likelihood of specific focal cranial nerve deficits increases. Many schwannomas exhibit combinations of intra- and extracranial effects. Because of these observations, it is useful to consider schwannomas not only according to the nerve of origin, but
519
also according to the anatomic compartments in which they occur and subsequently grow.
Specific Anatomic Sites By far the most common cranial nerves of origin for nonvestibular skull base schwannomas are the trigeminal (CN V), the facial (CN VII), and the jugular foramen group (CN IX, X, and XI). Therefore, these sites will be discussed in detail before addressing the other cranial nerve sites and noncranial nerve sites of origin, which are rare.
Trigeminal Nerve Schwannomas The trigeminal nerve is the second most common site for skull base schwannomas, after the vestibular nerve. The precise incidence is not known, but Kouyialis et al. in 2007 stated that a literature review dating back to 1935 turned up at least 580 published cases of trigeminal nerve schwannomas (TNSs) that had been surgically treated; undoubtedly many more were unreported or not treated surgically (32). Symptoms vary with exact site of origin, and manifestations are many, reflecting the long course and rich distribution of the trigeminal nerve. Patients with TNS present most often with trigeminal nerve-related dysfunction, including facial pain, headache, and numbness of the affected nerve segment(s). The character of symptoms varies considerably. In some patients pain is predominant and may be sharp or dull, intermittent or constant. In others, sensory disturbance is the chief concern. Patients often describe a sensation of numbness, burning, creeping, pins-and-needles, or other vague dysesthesias. Initially symptoms may be focal, but as tumor grows, deficits may become manifest in all three divisions. TNSs may develop essentially anywhere along the course of the trigeminal nerves, including the root, the ganglion, and the peripheral branches. According to Eisenman, the majority of trigeminal schwannomas develop in the gasserian ganglion (33), where they expand gradually from Meckel’s cave and initially involve only the middle fossa. With further enlargement, they may extend into the posterior fossa secondarily, with one component or the other being larger, creating a dumbbell shape [Figs. 7 and 2(B)]. Petrous apex erosion is common with such tumors, usually preserving the internal auditory canal, a finding which helps distinguish them from vestibular schwannomas (34).
Figure 7 Axial MR of patient with large trigeminal schwannoma involving posterior fossa (with brainstem compression) and middle fossa (at region of gasserian ganglion).
520
Nuss and Burke
and width of petrous erosion, classifying tumors as noted below (37): Type M, tumor confined to middle fossa Type Mp, tumor mainly in middle fossa Type M = P, tumor equally distributed into both middle and posterior fossae Type Pm, tumor mainly in cisternal space of posterior fossa
Management of Trigeminal Nerve Schwannomas
Figure 8 Coronal CT of giant V3 schwannoma extensively involving middle cranial fossa and infratemporal fossa. The patient reported only minimal numbness and no other symptoms.
TNSs may also arise directly from the trigeminal root near the brainstem in the posterior fossa. When tumor is confined to the posterior fossa, symptoms may mimic cerebellopontine angle tumors, causing hearing loss, vertigo, tinnitus, and facial weakness, in addition to trigeminal dysfunction, but in such cases the trigeminal dysfunction usually overshadows other deficits. Still other TNSs will arise more distally, and may even be entirely extracranial, in which case symptoms are very site-specific. For example, V1 tumors may arise either in the cavernous sinus or in the orbit, where they cause diplopia and proptosis due to pressure; the corneal reflex may be absent. They may also arise in V2, causing midfacial or palatal numbness, pain, and dysesthesias. Some patients may have xerophthalmia due to reduced lacrimation. V3 tumors result in numbness, pain, and dysesthesias of the lower face and jaw as well as chewing problems, malocclusion, and atrophy. Such tumors may become quite large as they extend into the infratemporal fossa, where there is ample room for slow growth before symptoms appear. Typically slow-growing, they may reach impressive proportions before some patients perceive symptoms (Fig. 8). Distal trigeminal schwannomas may also present in the frontal, ethmoid, sphenoid or maxillary sinuses, pterygopalatine fossa, or even in subcutaneous or submucosal tissues. Rarely, TNSs may involve more than two compartments, typically middle fossa, posterior fossa, and infratemporal fossa simultaneously (32).
Radiographic Appearance of Trigeminal Schwannomas A variety of radiographic appearances are possible for TNS, depending on the site of origin and direction of growth (Figs. 2, 5–8).
Staging of Trigeminal Schwannomas The generally accepted staging system used for TNS is that adopted by Samii et al. (35,36), summarized here: Type A: tumor predominantly middle fossa Type B: tumor predominantly posterior fossa Type C: tumor in both middle and posterior fossae Type D: tumor predominantly extracranial Gwak et al. proposed a different staging system, focusing on degree of petrous erosion, and felt that their system helped predict which surgical approach would be most successful. They used CT and MR to determine tumor diameter
Historically, TNS management has mostly been surgical, which is understandable since many patients initially present with symptoms caused by tumor compressing vital structures. Reports of successful radiation treatment are increasing in frequency, and will be discussed in the section on radiation. Surgical approaches are determined on the basis of tumor location and extent. A myriad of techniques and approaches have been advocated in the literature. The generally accepted approaches are as follows: Type A tumors are approached via temporal craniotomy, often with zygomatic or orbitozygomatic osteotomy for more basal access (Fig. 9). Type B tumors are often approached via suboccipital craniotomy or variations thereof. Type C tumors are best treated using combined middle-fossa/posterior fossa approaches such as combined subtemporal and presigmoid craniotomy. Successful one-stage resection can be accomplished (38–40), and appears to be becoming the standard of care, but some of these larger tumors may be best managed with two-stage operations, and treatment must be individualized. Type D tumors that are infratemporal are usually removed via infratemporal fossa approaches (33). Those Type D tumors that present in other areas, such as the pterygopalatine fossa, orbit, or sinuses, can be approached using transfacial, degloving, endoscopic transnasal, orbitotomy approaches, or combinations thereof (41–46). Details of surgical technique are reviewed elsewhere in this text.
Outcomes for Surgical Treatment of Trigeminal Nerve Schwannomas As in other areas of skull base surgery, major advances have been realized in the past two decades for TNS. Zhou et al. published their single-institution series of 57 cases of dumbbellshaped schwannomas in 2007 (39). Their analysis was divided into two groups, one treated pre-1984 using traditional (“non-skull base”) approaches, and the other after 1984, using modern microsurgical combined approaches. The two groups differed substantially in outcomes, with the pre-1984 cohort faring worse in all respects, most notably likelihood of total tumor resection (42% vs. 87%); long-term cranial nerve morbidity (55% vs. 18.6%); recurrences (3 of 12 vs. 1 of 45); and overall performance status. The authors concluded that the best treatment for large or giant TNS is microsurgery, via single-stage skull-base craniotomy (using a technique which they called an extraduro-transduro-transtrigeminal pore approach). The authors felt it was not necessary to resect the petrous apex for removal of the tumor in the posterior fossa. They recommended that radiosurgery be reserved for residual or recurrent tumors. Al-Mefty et al. described experience with 25 large dumbbell TNS in 2002, emphasizing a philosophy of singlestage, total tumor resection via an extradural zygomatic middle fossa approach through the expanded Meckel cave (40). All tumors involved the cavernous sinus, and most had more than one preoperative cranial nerve deficit, including abducens paralysis in 40%. In patients who had not undergone previous surgery, the preoperative trigeminal sensory
Chapter 37: Schwannomas of the Skull Base
521
Figure 9 Overview of orbitozygomatic-temporal approach as would be used for Type A trigeminal schwannoma. (A) Exposure for temporal craniotomy. (B) Orbitozygomatic osteotomies to facilitate infratemporal exposure. (C) Tumor (arrows) in infratemporal fossa seen protruding from foramen ovale before craniotomy [same orientation as in 9(A)]. (D) Preoperative (left) and postoperative MR images, demonstrating complete resection. (E) Appearance of patient approximately six months after surgery. Temporal fossa depression resulted from the use of the temporalis muscle as a reconstructive flap, and was not due to atrophy; patient declined temporal fossa reconstruction and remains tumor-free more than five years later.
522
Nuss and Burke
deficit improved in 44%, facial pain decreased in 73%, and trigeminal motor deficit improved in 80%. Among patients with preoperative abducent nerve paresis, recovery was attained in 63%. Only three patients (12%) experienced a persistent new or worse cranial nerve deficit, all confined to the trigeminal, postoperatively. Three patients experienced recurrences. The authors noted that, as with patients who have vestibular schwannomas, advances in surgical procedures have markedly improved outcomes in patients with trigeminal schwannomas, and that “preservation or improvement of cranial nerve function can be achieved through total removal of trigeminal schwannoma” (40). In a similar study that included not only 13 cases of TNS but also 27 other cases of benign nonmeningeal cavernous sinus tumors (47), improved cranial nerve outcomes were also noted. Postoperatively, 89.7% of the patients had either stable or improved extraocular muscle function compared with their preoperative statuses. Forty percent of the patients experienced improvement of their preoperative extraocular muscle deficits. From this experience, it was felt that with microsurgical skull base techniques, benign nonmeningeal tumors of the cavernous sinus, including schwannomas, can be safely and radically removed, and with good long-term neuro-ophthalmological function and low morbidity. In 2007, Pamir et al. reviewed a similar experience with 18 cases of TNS treated since 1992. Total excision was achieved in 17 of 18 cases, with minimal morbidity (36). Schwannomas affecting the orbit present special problems. TNS involving V1 and sometimes V2 can directly involve the orbit, with ocular complaints as the presenting symptoms. A wide spectrum of clinical findings can be observed (48), but usually progressive proptosis is the earliest sign (49). Vision is not usually impacted early on, but direct optic nerve compression, globe indentation with induced hyperopia, or increased intracranial pressure with optic nerve compromise may be responsible for visual decline as tumors progress. Careful documentation of neuro-ophthalmological findings is essential in all patients with trigeminal schwannomas. The ophthalmologist must play an active role in the multidisciplinary team to ensure optimal assessment, treatment planning, and visual outcomes, including rehabilitation of gaze deficits. In surgical management of TNS, the philosophy of total tumor resection appears to be extremely important. Multiple reviews have noted that when gross total tumor removal has not been accomplished, TNS tumors are highly likely to recur. In Gwak’s series of 29 consecutive TNSs, recurrence was noted in 8 of 10 patients in whom total resection was not achieved (37). Moffat et al. reported a series of eight TNS patients, among whom they elected to do subtotal resection in five (50). In those cases, tumor was left behind with the goal of minimizing postoperative deficit and improving quality of life. All five recurred and required revision surgery. Surgical resection remains the mainstay of treatment for trigeminal schwannomas. With advances in microsurgical techniques, creative skull base approaches, endoscopy and related technological improvements, morbidity of resection is decreasing, and the likelihood of improvement of preoperative neurological deficits is increasing. The role of radiation will be discussed later in this chapter.
Facial Nerve Schwannomas Facial nerve schwannomas (FNS) are the next most frequent after trigeminal nerve schwannomas, and are also uncommon. A literature review in 1996 revealed just over 300 published cases (51), but since that time over a hundred new cases have been added, probably reflecting increasing diagnostic
accuracy and awareness. FNS may occur anywhere along the course of the facial nerve, and like schwannomas elsewhere, symptoms are dependent upon which area is involved and to what extent the tumor has grown. Facial paralysis is the eventual hallmark of these tumors and a major source of morbidity among affected patients. Like the trigeminal nerve, the facial travels a long course and is anatomically complex in its distribution. Dort and Fisch (52) classified FNS tumors as intracranial, intratemporal, and extratemporal, with symptoms being different for each. Intracranial lesions (i.e., at the cerebellopontine angle) and those in the internal auditory canal frequently result in sensorineural hearing loss, tinnitus, and vestibular symptoms because of compressive effects on the eighth nerve. Intratemporal tumors often present with facial palsy and with conductive hearing loss, which occurs as tumor expands into the middle ear space. Extracranial tumors basically present as parotid or retromandibular masses at the skull base. Schaitkin authored a scholarly review of FNS in 2000 (53), and the reader is referred to that source for a detailed historical summary of reported cases of FNS. In his review, and those of O’Donaghue and Wiggins (54,55), the frequency of site of origin was examined. The majority of FNS tumors originate in the region of the geniculate ganglion (up to 83%), followed by the labyrinthine and tympanic segments of the facial canal (54% for both). Nearly 30% of these tumors will cause erosion of the otic capsule, visible on CT. Very few FNSs appear to originate in the CP angle, unlike vestibular schwannomas, but when they do they may be indistinguishable from VS both clinically and radiographically. Schaitkin observed that multiple segments of the facial nerve are commonly involved at diagnosis, most likely because FNS tumors are so indolent that they become extensive before they cause symptoms. In fact, multiple authors have noted that FNSs are extremely slow-growing tumors (56,57), in some cases with many years transpiring before definitive diagnosis is made. A significant number of FNSs have been noted at autopsy without ever having caused clinical symptoms (53,58). Lipkin reviewed 238 previously reported cases of FNS, and many of the findings were similar to those of schwannomas at other sites (59). The mean age at diagnosis was 39 years; there was no gender predilection; there was no distinct laterality of incidence; and the exact location of the tumor was the best predictor of symptomatology.
Clinical Features of Facial Nerve Schwannomas The most common clinical presentation of FNS is that of slowly progressive facial paralysis, which may sometimes be preceded by facial twitching, spasm, tic, or pain. This evolution is in contradistinction to the classic Bell’s palsy, which presents as facial palsy of sudden onset and rapid progress. That said, however, in several of the FNS series, a minority of patients did not have slowly progressive paralysis but instead had either sudden complete paralysis (14–21%) or recurring ipsilateral paralysis (up to 10%) (53,59,60). Thus, facial nerve tumors may mimic Bell’s palsy in a minority of cases, and clinicians must be cognizant and thorough in evaluation and follow-up of such patients. Otologic symptoms including hearing loss, tinnitus, and disequilibrium may be prominent in some patients with FNS. Although facial nerve symptoms usually occur first, Lipkin found that up to 13% had tinnitus initially. McMenomy reported on 12 FNS patients in whom 100% had hearing loss, 50% had tinnitus, and none had facial nerve symptoms, essentially presenting with symptoms indistinguishable from
Chapter 37: Schwannomas of the Skull Base
523
praxia only, but paralysis has been of a slowly progressive nature.
Radiologic Characteristics of FNS
Figure 10 Imaging of facial schwannoma involving multiple segments of nerve. (A) Axial T1-weighted, contrast-enhanced MRI at the level of the leftside CPA. Left-side facial schwannoma is seen as an enhancing tumor mass involving the CPA, IAC, labyrinthine segment (arrowhead), and geniculate fossa (arrow). The overall shape of the tumor is that of a dumbbell. (B) Axial computed tomography scan filmed in bone window at the same level as in A. The widened labyrinthine segment of the facial nerve (arrowhead) and expanded geniculate fossa (arrow) are evident. Source: From Ref. 66.
vestibular schwannomas (61). (These differences in reported incidence may reflect practice/referral patterns more than actual incidence, but they are noteworthy nonetheless. The point is that FNS can mimic VS, Bell palsy, and other entities.) Park et al. (62) emphasized the diagnostic dilemma of preoperatively distinguishing FNS from VS, and analyzed multiple parameters including preoperative symptoms, pure-tone audiometry, auditory brainstem response, caloric test, electroneuronography, and magnetic resonance imaging; ultimately none could reliably predict FNS. Kubota et al. reported two cases of facial nerve schwannoma in which there was no paralysis despite massive tumor involving both middle and posterior fossa, with the only symptoms being otologic (63). The diagnostic dilemma applies as well to intraparotid FNS, where definitive preoperative diagnosis is rare. Fine needle biopsy of FNS is often inconclusive because these tumors are hypocellular or heterogeneous. There have traditionally been no universally accepted imaging criteria that favor FNS over other more common tumors of the gland. Recently, however, Shimizu et al. described the so-called “target sign” in which T2-weighted MRI images reveal higher signal intensity around the tumor periphery, which they believe is suggestive of schwannoma (64). Whether this will be of benefit in the prospective recognition of FNS is not yet clear. At present, most intraparotid schwannomas are not diagnosed definitively before surgery. Caughey suggested that schwannoma should be suspected if the facial nerve cannot be found intraoperatively or if the tumor is intimately associated with the facial nerve (65). Other regional signs and symptoms of FNS include 11% with pain in or around the ear, 13% with visible mass in the auditory canal (caused by tumor in vertical segment), and 6% with otorrhea. Occasional patients have reported disturbed taste and salivation when there is involvement of the chorda tympani. With schwannoma, as with other causes of facial weakness, electrophysiological testing may provide helpful clues. Schaitkin has recommended a protocol for facial paralysis evaluation that includes facial EMG and evoked EMG (EEMG) (53). He proposed that tumor should be suspected if (i) prolonged conduction latency is present on EEMG, even with normal amplitude; (ii) a patient has an incomplete facial lesion with normal EMG but EEMG amplitude less than 10%, and (iii) there is EEMG and EMG evidence of neuro-
Schwannomas of the facial nerve share the same general characteristics as schwannomas at other sites. As with trigeminal and jugular fossa schwannomas, it is essential to obtain highly detailed, thin-section images in three planes, including CT and MRI, with contrast. Typical findings include increased diameter of any segment of the facial nerve canal in a fusiform shape; sharply defined bone erosion around the geniculate ganglion, otic capsule, or other nerve canal segment; and middle ear mass, cerebellopontine angle mass, or parotid mass (55,65,66). Various imaging appearances are shown in Figures 10 to 12. Often, intraparotid FNS will be contiguous with tumor involving the vertical and sometimes horizontal segments of the facial nerve in the temporal bone, and the facial canal in the mastoid will be widened (Fig. 13).
Figure 11 Imaging of facial nerve schwannoma continued. (A) Axial CT scan at the level of the epitympanum. Left-side facial schwannoma involving the geniculate fossa (closed arrow) and tympanic segment of the facial nerve (arrowheads) can be seen pedunculating into the middle ear. The lateral displacement of ossicles by tumor (open arrow) is evident. (B) Axial T1weighted, contrast-enhanced MRI at the same level as in Fig. 4(A) shows the schwannoma as an avidly enhanced oval mass (arrow). Source: From Ref. 66.
524
Nuss and Burke
Figure 12 Imaging of facial nerve schwannoma continued. (A) Axial CT scan filmed in bone window at the level of the mid-mastoid segment of the facial nerve canal. The left-side mastoid segment facial schwannoma is seen as an oval mass breaking into surrounding mastoid air cells (arrow). The tumor also dehisces the posterior wall of the external auditory canal (arrowhead). When the schwannoma breaks into surrounding mastoid air cells, it gives the impression of irregular “invasive” tumor margins. (B) Coronal CT scan filmed in bone window shows the same tumor in its craniocaudal extent. The enlarged stylomastoid foramen (arrow) is easily identified from this vantage point. (C) T1-weighted, contrast-enhanced coronal MRI of the same tumor. The inferior limit of tumor enhancement is the stylomastoid foramen (arrowhead), differentiating the schwannoma from a primary parotid neoplasm with perineural spread. Source: From Ref. 66.
(A)
(B)
Figure 13 Imaging of facial nerve schwannoma continued. (A) Axial CT showing facial nerve schwannoma (arrows) presenting as parotid mass. (B) Coronal MR of the same patient showing tumor (fusiform area of high signal intensity) involving vertical segment of facial nerve in mastoid.
Chapter 37: Schwannomas of the Skull Base
525
Figure 14 Topographic classification of facial nerve schwannomas as proposed by Litre et al. Images are obtained from a fused bone window computed tomographic scan with magnetic resonance imaging sequences. The tumor margin is delimited by the outline. Source: From Ref. 67.
Litre et al. described a classification for facial nerve schwannoma, based on a fused CT-MR algorithm (67) (Fig. 14) designed to aid in treatment planning. Classification of facial nerve schwannomas (67) Type I—tumor localized on geniculate ganglion. Type II—tumor is dumbbell-shaped on the geniculate ganglion, labyrinthine segment, internal auditory canal, and cerebellopontine angle cistern. Type III—tumor develops in tympanic and/or vertical segments of the facial nerve. Type IV—tumor develops in internal auditory canal or cerebellopontine angle without invasion of fallopian canal or geniculate ganglion. This category is difficult to distinguish from the vestibular schwannoma using radiological criteria. In this group, diagnosis was usually based on a previous microsurgical attempt.
regions. Intracranially, vestibular schwannoma may be indistinguishable; meningioma usually will show a “dural tail” sign, although that sign has been shown to be occasionally unreliable (68,69). In the temporal bone, cholesteatoma can cause similar bone erosion and facial paralysis. Traumatic neuroma, related to prior chronic otitis or temporal bone trauma, looks very similar to FNS. Granular cell tumor and osseous hemangioma are two entities that occur in the geniculate region. The former usually destroys bone irregularly, as opposed to the FNS. The latter is distinctive on CT, with “salt and pepper” density and a margin of new bone formation. Metastatic lesions to the temporal bone may also be considered, but the bone erosion is typically less well circumscribed. Extracranially, the chief differential diagnosis is that of parotid neoplasia.
Differential Diagnosis of Facial Nerve Schwannomas
Management and Outcomes for Facial Nerve Schwannomas
The differential diagnosis of FNS includes a variety of other common and uncommon pathologies of the same anatomic
As with trigeminal schwannomas, FNS has traditionally been managed surgically. Usually, achievement of gross total
526
Nuss and Burke
tumor resection has required resection of the nerve of origin. The rationale for aggressive total tumor removal, and nerve resection if necessary, is that as the tumor grows, there is progressive degeneration of axons and collagen deposition, which ultimately decrease the likelihood of good functional outcome (70). Therefore, better outcomes could theoretically be expected with earlier intervention. With FNS, the obvious concern is the facial paralysis that is inevitable after the facial nerve has been resected and reconstructed. Even in expert hands, nerve resection and grafting yields imperfect facial animation, and in most series the best outcome can be expected to result in function at House-Brackmann grade 3 or 4; many fare worse (57,71–73). While resection and nerve reconstruction are clearly necessary in patients who initially present with moderate or severe paralysis, management of patients with little or no paralysis is more challenging. In peripheral nerve surgery, it is often said that schwannomas can be separated from their nerve of origin and functional axons can be preserved intact. With schwannomas of the cranial nerves, this has not often proved to be possible in the senior author’s experience (DN), and in the reported experience of others (53,56,74). However, of all locations in which one would want to remove tumor without disrupting the affected nerve, the facial nerve would be a high priority for preservation. Conley and Janecka recommended that nerve preservation surgery for facial schwannoma “should be attempted every time,” but noted that in the majority it was not possible to preserve intact nerve. Of nine intratemporal FNS in their series, only one allowed preservation of the nerve (74). However, a number of subsequent authors have published increasing experience with nerve-preserving techniques, which can yield excellent functional outcomes in selected patients (57,70,75). In 2007 Lee et al. reviewed their experience with nervepreserving approaches for FNS (75). Using microsurgical technique and EMG monitoring, they employed what they termed “stripping surgery,” separating tumor from intact axons. The technique was used in six patients who had good preoperative facial function (HB Grade 1 or 2); it was possible to completely remove tumor in four. In the remaining two they performed decompressive surgery only (see below). Tumors were located in the geniculate ganglion (1 case), mastoid (2 cases), and IAC (3 cases). In all cases facial nerve integrity was preserved. Good facial function was preserved in all; two patients achieved HB Grade 1, four achieved Grade 2. There were no recurrences in the follow-up period, which ranged from 6 to 128 months (mean, 53 months). Size and location of tumor did not affect outcome. Their recommendation was that nerve-sparing approaches should be attempted in all patients with good preoperative facial function. In cases where tumor cannot be separated from intact nerve fibers, decompression may be a good alternative (75,76). Decompression applies to tumors that are expanding within the bony confines of the temporal bone. For patients with intratemporal FNS who have little or no preoperative paralysis, this technique simply removes bone from the areas around the tumor, with the goal of delaying compressive symptoms. Angeli and Brackmann reported excellent facial function in 4 of 4 patients in whom this technique was used and negligible or no tumor progression was noted during postoperative surveillance averaging 45 months. Interestingly, one of their patients presented with HB Grade 5 facial weakness, and improved to Grade 2 (75,76).
A wide variety of surgical approaches have been used to remove or decompress FNS, depending on tumor site and extent. Temporal bone approaches, middle fossa and posterior fossa approaches, and parotidectomy approaches have been used in various combinations; these are presented in detail elsewhere in this book. The complex subject of facial nerve reconstruction is relevant to this discussion but beyond the scope of this chapter.
Observation for Facial Nerve Schwannomas A cogent argument can be made for observation alone in selected FNS patients. Again, it is generally agreed that patients presenting with moderate or severe facial paralysis should undergo nerve resection and reconstruction, but for those with little or no paralysis, observation is an option. Proponents for observation (also referred to as conservative or expectant management) correctly point out that FNS is an extremely indolent tumor in most cases, perhaps more so than schwannomas at other sites, and that even when no therapy is given, patients may not experience progression of symptoms for years. Liu and Fagan (71) reported a series of 22 FNS cases, in which 12 were excised and 10 were observed. The best postoperative facial function in the tumor removal group was HB Grade 3, while 8 of the 10 conservatively treated patients had normal facial function up to 10 years later. No significant tumor growth was noted in any of the observed patients. The authors concluded that, for patients with little or no facial nerve symptoms, “delaying surgical resection of facial nerve schwannomas may allow patients to retain normal facial function indefinitely.” Not all observed patients will fare well, however, Perez et al. reviewed 24 FNS patients, of whom 11 underwent surgery and 13 were observed. Of the 11 in the operated group, 6 had unchanged postoperative facial function, 4 were improved, and one was worse; there were no postsurgical recurrences (57). Of the 13 in the observed group, facial function remained unchanged in 8 but was worse in 5 patients. In 4 patients, tumor progression was noted, and 3 underwent subsequent surgery. The authors concluded that “the decision on how to treat these patients should be individualized and based on initial facial function, growth rate, surgical experience, and informed patient consent.” A significant issue in considering the observation alternative is the establishment of a definitive diagnosis. Certainly, patients who present with classic findings of FNS may be candidates for observation without invasive diagnosis, but it must be acknowledged that the diagnosis can be elusive, and a close follow-up protocol must be strictly adhered to. As noted above, many FNS tumors are clinically and radiographically similar to other pathologic entities for which observation alone would be a poor choice. Therefore, a prudent approach would be to obtain a tissue diagnosis in cases where there is doubt. In all cases of FNS, the choice of “observation” should be viewed as a firm commitment to regular and detailed clinical and radiographic follow-up for the rest of the patient’s life. In addition to surgery and observation, stereotactic radiosurgery has received increasing attention in recent years. This will be reviewed in the section on radiation.
Decision-Making in Management of Facial Nerve Schwannomas Patients with small, minimally symptomatic FNS in whom the diagnosis is confirmed or considered highly likely may be offered several good alternatives, including nerve-preserving
Chapter 37: Schwannomas of the Skull Base
surgery, decompression if tumor is not separable, observation, or stereotactic radiation. Significant controversy remains as to which of those options is best. Patients with larger tumors and/or moderate to severe symptoms should undergo surgery to remove the tumor, and in most cases to resect and reconstruct the facial nerve. Radiosurgery may be considered as an alternative in selected cases, or as an adjunct for residual or recurrent tumor.
Jugular Foramen Schwannomas It is important to realize that while jugular foramen schwannomas are among the more commonly encountered nonvestibular schwannomas, they are still rare. The world’s literature on these tumors recorded only about 200 cases as of the year 2000, and most of what is written has come from single case reports and small series. Of the largest reported series of jugular foramen schwannomas, most describe experience with fewer than 20 patients. Therefore, what is known of these tumors is based on relatively limited information. An elegant and thorough review of this subject was presented by Von Doersten, who did a meta-analysis of all previously published cases in 2000 (77). He estimated that the incidence of jugular foramen schwannomas was likely in the range of 3to 5 tumors per 10 million people per year. Jugular foramen schwannomas (JFSs) originate from the nerves of the jugular foramen, namely cranial nerves IX, X, and XI. These tumors are generally grouped together for discussion purposes because it is sometimes difficult to precisely determine the exact nerve of origin, given the tight proximity of the three nerves in and around the jugular fossa, and also because of the tendency for tumors to involve more than one nerve simultaneously. The jugular foramen is a region of densely compacted, intricate anatomy. Basically, it is an intracranial/extracranial conduit from the posterior fossa to the parapharyngeal space and upper neck, delineated by a smooth opening between the occipital bone and the petrous portion of the temporal bone. Traditional surgical anatomy describes two parts, the pars nervosum anteromedially, and the pars venosum posterolaterally. In the pars nervosum, the nerves are situated such that CN IX is most anterolateral, and CN XI is most posteromedial, with the vagus in between. Surgical approaches to this compartment are challenging because the facial nerve, internal carotid artery, mastoid, middle ear, and inner ear structures are all within a few millimeters of the jugular foramen, and therefore are vulnerable to injury. For this very same reason, untreated tumors in this region can be treacherous and even lethal. A critical issue in management of these tumors is the status of the jugular vein. In the majority of persons the right jugular vein is larger, reflecting the dominant venous outflow from the posterior fossa; this is considered a normal finding. In some individuals the left vein will be larger. Generally if the vein is known to be completely occluded preoperatively, it is assumed that sacrifice will not be likely to create new problems postoperatively. However, it is interesting that these tumors, despite their expansion in a tight space, do not always lead to complete obstruction of the venous circulation. Injury to, or sacrifice of, a functioning jugular vein can lead to disastrous cerebral venous infarction. Symptoms and signs of JFSs are quite variable. When tumors are small they may be asymptomatic. As mentioned above, the pattern of growth determines what symptoms will evolve. Early extracranial growth results in typical deficits of CN IX, X, and XI, with dysphagia, hoarseness, and shoulder weakness, respectively. When all three nerves are affected, the
527
patient is said to have Vernet syndrome. When extracranial expansion exerts pressure on CN XII, the patient also develops tongue weakness and is said to have the Collet–Sicard syndrome. A variety of other eponyms have been applied to related symptom combinations of this region (77). When the tumor grows intracranially, the symptoms relate to expanding tumor causing pressure on nearby structures of the posterior fossa and cerebellopontine angle. In some of these tumors, the intracranial symptoms will predominate. Thus, hearing loss, vertigo, tinnitus, facial weakness, tongue weakness, and signs of elevated intracranial pressure (headache, hydrocephalus) develop. With progression, patients develop gait disturbance, ataxia, long tract signs including motor weakness in the extremities, and potentially may die as tumor compresses the brainstem. In some patients, these intracranial manifestations may be present without neuropathies of cranial nerves IX, X, or XI. Schwannoma cases have been reported in which the presenting sign was foramen magnum syndrome (78) (defined by unilateral arm sensory and later motor deficits, progressing to ipsilataeral leg, then contralateral leg, and finally contralateral upper extremity deficits) (79); schwannoma has also been reported to mimic tumor of the fourth ventricle (80). Typically, these tumors will grow in both directions to some extent. The path of least resistance allows bulbous expansion above and below the foramen, with bony erosion about the foramen, giving the classic dumbbell shape (77) (Fig. 15). Von Doersten compiled a database of 164 analyzable JFS cases from the existing literature (i.e. those reports that contained sufficient detail for meaningful review) (77). Of those cases, the mean age at diagnosis was 41.7 years, with equal gender distribution. The nerve of origin of the JFS was determined to be CN X in 43.5%, CN IX in 29%, and CN XI in 14.1%, with the remainder involving two or more of the nerves. Thus, in descending order of frequency, tumors appear to arise from the vagus, the glossopharyngeal, and the spinal accessory nerves, followed by combinations thereof. In Van Doersten’s review, the most common symptoms were hearing loss (35%), hoarseness (26%), dysphagia (12%), and vertigo (10%), with other less common complaints occurring in many combinations of both intracranial and extracranial symptoms. Depending on size and extent of tumor, cranial nerve deficits from V to XII were all reported, individually (except that there were no isolated presentations of sixth nerve palsy) and in various combinations. Clinical evaluation of patients with suspected lesions of the jugular foramen includes careful history and physical examination, with detailed neurological focus on cranial nerves, cerebellum, and long tract signs. Audiologic and vestibular studies quantify sensory deficits of the ear. Flexible endoscopy of the upper airway documents details of lower cranial nerve function. Inquiry must be made as to any history of chronic lung disease and gastroesophageal reflux disease, since these disorders may place the patient at higher risk of poor outcome from aspiration, which commonly occurs with JFSs. All of these details influence preoperative planning and patient counseling regarding the need for feeding tubes, laryngoplasty, facial reanimation procedures, and hearing and balance rehabilitation (77). Radiologic investigation of suspected jugular fossa tumors must be thorough. Thin-section CT and MR in all three planes are necessary to optimally evaluate these lesions; often MR arteriography and MR venography are necessary to adequately image the regional bloodflow. In addition to the
528
Nuss and Burke
Figure 15 Comparison of early and later stages in development of jugular foramen schwannomas in three planes (from top to bottom, parasagittal, axial, and coronal planes). Images on left side of page show small tumor within jugular foramen. Note that with small tumor, venous system remains patent. Images on right side show larger schwannoma, now bilobed and involving both intra- and extracranial compartments, with compromise of venous system. The degree of intra- versus extracranial growth is highly variable. Source: from Ref. 77.
general radiographic features described for schwannomas, JFSs will exhibit specific findings that relate to the unique regional anatomy. On CT, the jugular fossa will be smoothly enlarged, and tumor will be isodense with muscle. On MR, T1 images will typically show low signal intensity; T2 will show higher signal. MR can give much information about flow in the jugular vein, which gives a dark signal void in high-flow states and a bright or mixed signal in low-flow states, but MR venography and arteriography can more accurately define regional circulation. On both CT and MR, dumbbell shape of the lesion is considered to be virtually pathognomonic for schwannoma, and the tumor will enhance brightly with contrast due to vascularity. Figures 16 to 18 show common radiographic appearances of JFS.
Differential Diagnosis of Jugular Foramen Schwannomas The differential diagnosis of lesions at the jugular foramen is shown in Table 1 (77). Paraganglioma is differentiated from JFS by a more irregular pattern of bone destruction on CT, and large flow-voids within the tumor, along with the classic “salt and pepper” appearance on MR. Meningioma will often be associated with the “dural tail” sign. Appropriate imaging will sometimes include traditional angiography to better define vascularity and regional bloodflow, and to help differentiate the likely tumor type. JFSs do not usually show an angiographic “blush,” typical of the much more vascular paragangliomas. Embolization can be helpful to preoperatively occlude the inferior petrosal sinus, which can be a troublesome source of bleeding
Chapter 37: Schwannomas of the Skull Base
529
Figure 16 Radiographic appearance of jugular foramen schwannomas. (A) Axial CT scan with bone algorithm shows an enlarged right jugular foramen (arrows). Note the sharp, rounded bone borders of the intraosseous extension, including a thin sclerotic rim and slightly bulging and eroded cortex (arrowheads). (B) Coronal contrast-enhanced T1-weighted MR image shows dumbbell-shaped tumor extending both into the posterior cranial fossa (arrows) and below the skull base (star). (C) Axial CT scan with bone algorithm shows a flared IAC meatus (arrow), which may be normal, although it is probably due in part to tumor eroding its posterior margin (compare with D). (D) Axial T2-weighted MR image shows a large tumor with high, slightly inhomogeneous signal in the posterior cranial fossa abutting but not extending into the right IAC (arrow). Note deformity of the brain stem, fourth ventricle, and cerebellum by the tumor. Source: From Ref. 28.
when the tumor is resected from the jugular bulb. Minimizing this bleeding may help limit injury to uninvolved cranial nerves (77). Despite the rarity of JFS several staging systems have been described. Kaye advocated a simple system in which stage A indicates tumor that is mostly intracranial; stage B mainly involves the bone surrounding the foramen; stage C is mostly extracranial (81) Pellet (82) added a stage D, defined as dumbbell tumor with both intra- and extracranial extent. Franklin (83) adapted the paraganglioma staging system of Fisch, describing four stages (and multiple substages of each), based on extension from the neck to the petrous carotid to the intracranial space, as outlined below: A: B: C: D:
tumor only in the neck tumor extends to jugular fossa but is primarily extracranial tumor progresses along petrous carotid tumor extends intracranially
Treatment Options for Jugular Foramen Schwannomas As with schwannomas elsewhere, the treatment options for JFS include surgery, radiation, and observation. Decision
making is complex for several reasons. First, the natural history for JFS, in contrast to the much more common vestibular schwannoma (VS), is not well known. With increasing access to imaging, these tumors are sometimes diagnosed in minimally symptomatic persons, or asymptomatic persons undergoing imaging studies for unrelated causes. There are insufficient studies of JFS upon which to advise such patients. Von Doersten postulated that if one were to extrapolate what is known from the abundant natural history studies of untreated VS, it would be reasonable to assume that JFSs can be expected to grow 0.1 to 0.2 cm per year. If that assumption is true, then very small, manageable tumors identified in young healthy persons would likely progress by 1to 2 cm per decade, eventually becoming formidable lesions. Theoretically, such tumors could be removed with less morbidity while small. However, observation might also be a valid option for patients with small tumors and minimal symptoms. Observation, or so-called “expectant management,” will be addressed in more detail later in this chapter.
530
Nuss and Burke
Figure 17 Radiographic appearance of jugular foramen schwannomas, continued. 21-year-old woman with otalgia and sudden seventh cranial nerve paralysis caused by jugular foramen schwannoma. (A) Axial CT scan with bone algorithm shows marked tumor growth eroding the bone (arrows). Note rounded, sharply demarcated margins as well as erosion and bulging of cortex along the medial border (arrowheads). (B) Coronal CT scan with bone algorithm shows marked intraosseous tumor growth, rounded and sharp bone borders, and erosion and slight bulging of cortex (arrows). (C) Sagittal noncontrast T1-weighted MR image shows the patient’s normal right side (arrow points to jugular foramen). (D) Sagittal noncontrast T1-weighted MR image shows a tumor with low signal intensity in the left jugular foramen, extending below the skull base (arrows). (E) Axial contrast-enhanced T1-weighted MR image shows tumor growth into the IAC (arrow), which was confirmed on coronal MR images (not shown). (F) Axial contrast-enhanced T1-weighted MR image shows that the main portion of the tumor is located within the jugular foramen and below the skull base (arrows) (compare with D). Source: From Ref. 28.
With larger tumors that are causing deficits, the decision to treat is not as difficult. Hence, the decision must focus on surgery versus radiation.
Surgical Treatment of Jugular Foramen Schwannomas In deciding how best to manage JFSs, outcomes of reported series managed surgically must be thoughtfully considered. In Von Doersten’s review, in which all cases were treated surgically with curative intent, it was noted that many papers did not specifically record preoperative cranial nerve status; some did not record postoperative status either. He stated that “it was very difficult to interpret whether any preoperative cranial nerves actually improved after resection of the tumor, or whether the status simply was not recorded.” (Intuitively, it seems that most surgeons would make note of clinical improvement in such reports if improvement did in fact occur.) Von Doersten also noted that in cases where CN IX, X, and XI deficits were noted preoperatively, there was no recovery of any cranial nerve function postoperatively. In
his review, cranial nerve function was possible but not likely when multiple deficits were present preoperatively. Moreover, the likelihood that additional (i.e., new) cranial nerve deficits would be caused by surgery was significant. Again, the inhomogeneity of the reports made interpretation difficult, but a few observations were insightful. Of the 31 cases in which little or no facial nerve deficit was present postoperatively (House-Brackmann score of 1 or 2) (84), none had any preoperative deficit. There were 10 patients with HB scores of 3 or 4, and only five of those had any preoperative deficit. Thus, all of the patients with the best facial nerve outcomes were those who had no preoperative deficit, and among patients with worse outcomes there were some who had normal facial function preoperative. Surgical approaches to jugular foramen tumors are numerous and addressed elsewhere in this text. The most frequently employed methods have included modifications of the Fisch infratemporal approach, suboccipital approach, petro-occipital trans-sigmoid (POTS) approach, retrosigmoid
Chapter 37: Schwannomas of the Skull Base
Figure 18 Radiographic appearance of jugular foramen schwannomas, continued. 63-year-old woman with jugular foramen schwannoma causing hearing loss, reduced gag reflex, and nystagmus. (A) Axial T2-weighted MR image shows tumor with high signal intensity within the enlarged left jugular foramen. The tumor is well demarcated with smooth borders. There is only a small tumor bulge into the posterior cranial fossa (arrow). (B) Coronal contrast-enhanced T1-weighted MR image with fat suppression shows strong and homogeneous contrast enhancement in schwannoma located below the skull base (arrow indicates level of the jugular foramen). Source: From Ref. 28.
approach, and combinations thereof. Unfortunately, nomenclature in existing reports is not uniform, making comparisons difficult. Tumor outcomes, reported for 164 cases, were such that only seven recurrences were documented. Of these, five had been operated via a suboccipital approach, one via a cervical approach, and one by an approach not documented. No recurrences were reported after infratemporal fossa resection. The conclusion was that combined procedures that afford access to both intracranial and extracranial aspects of the jugular fossa, above and below the tumor, are more likely to be curative. In a more recent report, Wilson et al. (85) reviewed their experience with seven JFS patients, of whom six were managed surgically. Patients ranged from ages 24 to 69 years. Presenting symptoms included dizziness, hearing loss, dysphagia, diplopia, tongue paresis, and hoarseness. All operated patients had complete tumor excision. Lower cranial nerve Table 1 Differential Diagnosis of Tumors of the Jugular Foramen Primary tumors r Paraganglioma r Meningioma r Schwannoma, neurofibroma r Hemangiopericytoma r Chondrosarcoma r Plasmacytoma Secondary tumors r Endolymphatic sac tumor r Nasopharyngeal carcinoma r Malignant tumors of the temporal bone (SCCA) r Parotid neoplasms r Langerhans cell histiocytosis Metastatic disease r Squamous cell carcinoma r Breast cancer r Prostate cancer Source: Modified from Ref. 77.
531
dysfunction that was present preoperatively did not resolve, whereas preoperative deficits of CN V and VI did resolve. The incidence of new lower cranial nerve deficits was 15%, and these were only in the nerves that were determined to be the nerve of origin. No recurrences were seen. In two cases, temporary feeding tubes were needed. The authors concluded that JFS tumors can be successfully managed with surgery, and with low morbidity. In 2006, Sanna et al. reported one of the world’s largest series of JFS patients, 23 cases treated over 18 years (86), in which 22 patients were treated surgically. A wide variety of approaches were used in various combinations, most often the petro-occipital trans-sigmoid approach (POTS), plus subtotal petrosectomy, translabyrinthine, transotic, and transcervical approaches, and infratemporal fossa type A approach, among others. Some required two-staged surgeries to address intradural and extradural tumor separately. Complete resection was accomplished in 21 patients, and there were no recurrences. There were no deaths, and only one CSF leak. Facial nerve function was preserved in all patients operated by the POTS approach, but a few patients operated by other approaches experienced HB scores of 3 and 4. Of patients who were operated with the intent of hearing preservation, good hearing was preserved in 83.3%. No patients recovered the function of the preoperatively paralyzed lower cranial nerves. A new deficit of one or more of the lower cranial nerves was recorded in 50% of cases. There was a single CSF leak and no perioperative mortality. The authors concluded that “surgical resection is the treatment of choice for JFS,” and that “the POTS approach allowed single-stage, total tumor removal with preservation of the facial nerve and of the middle and inner ear functions in the majority of cases. Despite the advances in skull base surgery, new postoperative lower cranial nerve deficits still represent a challenge.” A 2004 report from Kadri and Al-Mefty reviewed the experiences of six patients with JFS, who all presented with intra- and extracranial extensions through an enlarged jugular foramen (classic dumbbell tumors). Symptoms included deficits of CN IX, X, XII, and XI, in that order (87). All patients had two or more deficits upon presentation. A transcondylar suprajugular approach was used without sacrificing the labyrinth or the integrity of the jugular bulb. Complete resection was accomplished in all. There were no deaths, no new cranial nerve deficits, and no recurrences. There was one case of aspiration pneumonia. Two patients with preoperative deficits of CN IX and X improved and three patients recovered tongue mobility. Their conclusion from this series was that “with careful, extensive preoperative evaluation and appropriate planning of the surgical approach, dumbbell-shaped jugular foramen schwannomas can be radically and safely resected without creating additional neurological deficits. Furthermore, recovery of function in the affected cranial nerves can be expected.” To put these differing outcomes into perspective, and to make fair comparisons, a few observations are relevant. The “meta-series” reviewed by Von Doersten represents a great number of cases scattered across numerous centers worldwide with no uniformity of surgical approach, treatment philosophy, operating surgeons, or reporting strategies. It also spans many decades, across which skull base techniques, imaging studies, and instrumentation have changed dramatically. The series of Sanna et al. (among the most respected skull base surgeons of the world) is one of the largest series in the literature, but it shares some of the same characteristics,
532
Nuss and Burke
namely a diverse group of tumors, treated over a long timeframe spanning (and contributing to) the evolution of many techniques. The series of Wilson et al., and also the series of Kadri and Al-Mefty, are both smaller series but may actually be more relevant in some ways, because they represent more closely the current state of the art in skull base surgery, having taken place entirely in the “modern era of skull base surgery.” In Wilson’s series, the incidence of new cranial nerve deficit was just 15%, and all of those were confined to nerves of tumor origin. Essentially there were no unexpected or avoidable new palsies. Furthermore, associated upper cranial nerve deficits resolved. In Kadri and Al-Mefty’s series, unlike the others, improvement of lower cranial nerve deficits was observed postoperatively. The recent trends toward improvement in functional outcomes are encouraging. It should be emphasized, however, that these outcomes were reported by specialized skull base teams whose senior surgeons are vastly experienced. The rarity of JFS tumors mandates that if surgical treatment is to be recommended, it must be rendered by such teams.
Decision-Making in Management of Jugular Foramen Schwannomas As a general rule, surgical treatment for JFS is arguably the optimal treatment in younger patients or those who are symptomatic with progressive deficits, and who are reasonably good surgical candidates. With symptomatic patients who are elderly, or those who are medically infirm or otherwise poor surgical candidates, the best treatment option is probably radiation, which will be discussed in more detail later in the chapter. JFS patients, who are minimally symptomatic with small tumors, or asymptomatic, can be managed expectantly (observed) with serial imaging and clinical examinations.
Schwannomas of Other Skull Base Sites Olfactory and Optic Nerve Schwannomas: Controversy The pathological literature and the surgical literature are at odds as to whether olfactory and optic nerve schwannomas even exist. In traditional anatomical and pathological terms (1,5), these nerves are devoid of Schwann cell sheaths; therefore, theoretically they should never give rise to schwannomas. Despite this scientific assertion, rare cases of schwannoma involving the olfactory (88,89) and optic (90,91) nerves have been apparently well documented. This discrepancy might be explained by the recent characterization of specialized glial cells known as “olfactory ensheathing cells” (and presumably optic ensheathing cells exist as well), which share many features in common with Schwann cells, including neural crest origin as well as a number of molecular markers (92,93). When such cells become neoplastic, they may well be indistinguishable from Schwann cells. These are of course extremely rare tumors (94,95) Presenting signs include hyposmia and visual blurring.
Schwannomas of the Nerves of Ocular Motility Orbital schwannomas most often arise from the trigeminal nerve (49), but rare cases of oculomotor (96–98) trochlear (99–100), and abducens origin (101–103) have been welldocumented. These lesions may arise anywhere along the course of these nerves and may therefore be intracisternal (where they cause brainstem effects), intracavernous (causing diplopia), intraorbital (causing diplopia and proptosis) (46), or may occupy more than one compartment.
Schwannomas of the Hypoglossal Nerve The hypoglossal nerve is another rare site for schwannoma, but one in which substantial morbidity ensues (104–106). Only 26 cases of transdural (dumbbell) hypoglossal schwannomas are reported, and their management is challenging for many of the same reasons that jugular foramen lesions are challenging. Typical symptoms are dysarthria, dysphagia, and symptoms of brainstem compression. Treatment may require combined microsurgical approaches to decompress the brainstem, which is frequently affected, and adjuvant stereotactic radiosurgery.
Unusual and Curious Manifestations of Schwannomas Schwannomas, as we have noted, may occur in virtually any nerve of the head and neck. The most common sites and presentations have been reviewed above. To complete the picture, we consider here some examples of schwannomas that have been reported in the most unusual locations, or those causing unusual symptoms and manifestations. Halefoglu et al. reported a severe case of NF2 in which the unfortunate patient had synchronous schwannomas involving both hypoglossal nerves, both vestibular nerves, the right trigeminal, the left oculomotor, and the right abducens (107). Cheong et al. observed bilateral vidian nerve schwannomas, associated with facial palsy, which resolved after transnasal resection of the tumors (108). Schwannomas have also been reported originating from the greater petrosal nerve (109), Jacobson nerve (110) and the sympathetic plexus of nerves about the cavernous segment of the internal carotid artery (111). In the latter case, microsurgical resection resulted in excellent outcome except for Horner syndrome. Kamel at al. reported a vagal schwannoma of the cerebellomedullary cistern causing severe refractory neurogenic hypertension (112), believed to be secondary to compression of medullary centers that coordinate sympathetic control of blood pressure. Jagetia et al. described a case of a dumbbellshaped trigeminal schwannoma in which the patient’s predominant symptom was pathological laughter, which resolved immediately after surgical resection (113). The authors opined that this outcome supports the theory that the brainstem and perhaps the medial temporal lobe play some role in the control of spontaneous laughter. Hsu et al. studied eight cases of pathology-proven orbital schwannomas, and four were found to have associated pneumosinus dilatans affecting paranasal sinuses adjacent to the orbit (114).
RADIATION IN THE TREATMENT OF SKULL BASE SCHWANNOMAS Historically, surgery has been the treatment of choice for most patients with skull base schwannomas, and with good reason. For most of the 20th century, the only nonsurgical option was external beam radiation, with which it was impossible to discretely irradiate skull base tumors without delivering unwanted high doses to nearby uninvolved structures, which would result in unacceptably high incidence of neurologic injury. With increasing availability of stereotactic and imageguided radiation delivery systems in recent years, however, and the consequent ability to precisely deliver high doses of radiation to well-defined target areas, growing attention has been paid to radiation as both primary and adjuvant treatment for schwannomas. The vast majority of this experience has been with vestibular schwannomas, and incontrovertibly radiation is
Chapter 37: Schwannomas of the Skull Base
now a well-established alternative for treatment of VS. Although stereotactic radiation systems have been available for decades, clinical experience with the very uncommon nonvestibular schwannomas has until recently been more limited, and clinical reports with adequate long-term follow-up were insufficient. A significant and growing body of new literature has now begun to improve our understanding of how radiation fits into the scheme of management of these benign tumors. The fundamentals of physics, radiobiology, and details of radiation delivery systems are available elsewhere and will not be addressed here. One important technical point, however, is that the vast majority of data regarding radiation of skull base schwannomas comes from experience with the Leksell Gamma Knife system and variations thereof, including linear accelerator-based units. The term stereotactic radiosurgery refers to these kinds of extremely precise delivery systems. While other radiation delivery systems (e.g., fractionated radiation therapy and proton beam radiation) have been used to treat schwannomas, outcomes data for those are more limited. Therefore, this section will review reported clinical outcomes regarding the effectiveness, limitations, and risks of stereotactic radiosurgery in the treatment of skull base schwannomas. In evaluating the role of any treatment, the critical issues are two: efficacy and safety. Efficacy is best measured by the degree of symptom relief and tumor control. Safety is assessed by noting whether existing deficits or symptoms worsen; whether any new deficits are incurred as a result of treatment; by the appearance of early or late complications; and by the implications for any needed future treatment. The following sections will attempt to examine these issues.
Radiation for Nonvestibular Schwannoma in General In 2006 Flickinger and Barker reviewed a broad experience with radiosurgery for cranial nerve schwannomas (115). They analyzed outcomes from published series involving thousands of patients worldwide, most of whom were treated for vestibular schwannoma, but focusing on the issues of tumor control rates, dosing, and morbidity of adjacent cranial nerves. Although this study involved predominantly vestibular schwannoma patients, the data are relevant here with respect to functional outcomes for cranial nerves in proximity to tumor. The authors concluded that “the low morbidity and high long-term tumor control rates with radiation treatment have made it the choice of many patients who opt for active initial management for small- or medium-sized cranial nerve schwannomas.” Pollock et al. reviewed the Mayo Clinic experience with 23 patients treated with radiosurgery between 1992 and 2000 for a variety of nonvestibular skull base schwannomas of various cranial nerves, including the trigeminal (n = 10), jugular foramen group (n = 10), hypoglossal (n = 2), and trochlear (n = 1) (116). Nine of these had undergone prior surgery. With a median follow-up of 43 months, 22 of 23 tumors were smaller (n = 12) or unchanged in size (n = 10). The only tumor that failed to respond was the one “malignant schwannoma” included in the series. Four patients (17%) suffered radiation-related morbidity, including three with trigeminal tumors who suffered new or worsened trigeminal dysfunction. One patient developed Eustachian tube dysfunction after treatment for hypoglossal schwannoma. Of note, however, no patient with lower cranial nerve schwannoma developed any treatment-related hearing loss, facial palsy, or dysphagia after therapy. The authors pointed out that the high tumor control rates reported for vestibular schwanno-
533
mas could be expected to apply to the nonvestibular schwannoma (NVS) population, and asserted that compared to historical controls treated surgically, radiosurgery appears to result in less treatment-related morbidity, especially for tumors of the lower cranial nerves. Symptom relief in this series was not specifically detailed. The authors made comparisons to several reported surgical series that summarized results after microsurgical resection of NVS, and concluded that the radiosurgery treatment-related morbidity and tumor control rates compared favorably. Further, they asserted that patients with jugular fossa schwannomas did better with radiosurgery than with surgery, in terms of reduced cranial nerve morbidity. They did stipulate, however, that for patients with significant mass effect, or those with primarily cystic tumors, primary surgical resection should be the first choice.
Radiation for Trigeminal Nerve Schwannomas Sheehan et al. published their experience treating trigeminal nerve schwannomas (TNS) using radiosurgery in 26 patients, from 1989 to 2005 (117). The median follow-up was 48.5 months. Clinically, 18 patients improved (72%), four were stable (16%), and three were worse (12%). Imaging studies revealed tumor shrinkage in 12 patients (48%), no change in 10 (40%), and tumor growth in three (12%). They concluded that the risk/benefit ratio with radiosurgery was favorable for TNS patients, but that larger studies are needed to better evaluate long-term outcomes. Hasegawa et al. also examined outcomes after treatment of TNS with radiosurgery, in 37 cases, with a mean follow-up of 54 months (118). Clinically, 40% of patients had improvement in symptoms, but one patient worsened despite good radiographic tumor control. Radiographically, 20 patients (54%) showed tumor regression and 8 (22%) showed stable findings. In 5 patients (14%), however, tumor enlarged or uncontrollable facial pain developed with radiationinduced edema requiring surgical resection. As in Pollock’s series, the authors concluded that gamma knife was safe and effective for select patients, but that large tumors or those that are cystic or compressing brainstem or 4th ventricle should be treated with surgery as the first choice. Huang et al. reviewed outcomes of 16 TNS patients treated with gamma knife radiosurgery; six had prior surgery and 10 were treated primarily; mean follow-up was 44 months (119). Clinically, five patients improved and 11 were stable. Radiographically, the tumor control rate was 100%, with regression in nine and stability in seven. Significantly, there were no new cranial nerve deficits of any kind after treatment. The conclusion of the authors was that radiotherapy offered a reasonable alternative to microsurgery, either as primary or adjuvant treatment, which “controlled tumor growth, did not cause new deficits, and often improved presenting symptoms.” Pan et al. examined long-term results of radiosurgery for TNS in 56 patients, in one of the largest series to date (120). Fourteen had undergone prior surgery, 42 were treated primarily. Clinically, 14 patients had complete relief of symptoms (numbness or diplopia), and improvement of other deficits was seen in 25 patients. In 13 patients, trigeminal dysfunction either did not change or got slightly worse, and in 4 patients worsening symptoms were related to tumor progression. Radiographically, tumors disappeared in 7 patients, regressed in 41 patients, and were unchanged in 4 patients. Four patients experienced tumor progression, and one of those died 36 months following treatment. The overall tumor growth control rate was 93% (52 of 56 cases). The authors concluded that radiosurgery is effective for small and
534
Nuss and Burke
medium-sized TNS tumors, but larger tumors should be surgically removed or decompressed, particularly when in close proximity to the brainstem. Also on the subject of TNS, Peker et al. studied gamma knife treatment of 15 patients followed for a mean of 61 months post-treatment (121). As in other reports, the cohort included patients treated primarily as well as adjuvantly. As in Huang’s series, 100% radiographic control rate was noted, with regression in 13 and no change in 2. One patient developed transient facial numbness and diplopia. Again, the conclusion was that radiosurgery is associated with good tumor control and minimal risk of adverse effects. As with all forms of radiation, stereotactic radiosurgery has been associated with complications, duly noted in several of the above series. Akiyama et al. reported an unusual complication 15 months after gamma knife treatment, involving rapid tumor regrowth with extensive cyst formation and severe brainstem compression requiring urgent surgery (122). At operation, the trigeminal nerve had to be sacrificed because of dense pseudocapsule formation that was attributed to radiosurgery. The patient suffered chronic facial pain afterwards. The authors admonished that radiosurgery can induce fibrosis or degenerative change that complicates subsequent surgery.
Radiation for Facial Schwannomas Kida et al. reviewed the stereotactic radiosurgical treatment of 14 patients with facial nerve schwannoma (FNS) (123). Eleven of 14 presented with facial palsy; nine presented with hearing loss. After a mean follow-up of 31.4 months, facial nerve function was improved in five, stable in eight, and worse in one patient. None suffered any new hearing loss. One patient developed facial palsy immediately post-treatment that recovered to House-Brackmann Grade 3. Radiographically, 10 tumors regressed and 4 were stable (100% control rate). No tumors progressed. The authors concluded that radiosurgery should be the “treatment of first choice for facial schwannomas.” Litre et al. presented their experience in irradiating FNS in 11 patients (67). These were identified from a large population of 1783 patients with CP angle tumors treated in the same institution. Favorable previous experience with facial nerve outcomes among VS patients provided rationale for using radiosurgery to treat FNS. Mean follow-up was 39 months. Clinically, three patients with facial weakness improved, and none developed any new palsy or worsening of previous palsy. Radiographically, 10 were stable or regressed, but 1 required microsurgery due to cyst development. The authors proposed a classification of FNS into four anatomic subtypes stratified according to different clinical and surgical difficulties (Fig. 14). They felt that gamma knife radiosurgery could become “a first treatment option” for small or medium-sized FNS.
Radiation for Jugular Foramen Schwannomas Martin et al. examined outcomes for 34 patients with 35 jugular foramen schwannomas (one patient had bilateral tumors) treated with radiosurgery (124). Noting that JFS often presents with multiple lower cranial nerve deficits, and that surgical resection may be associated with significant morbidity, their hypothesis was that radiosurgery might reduce cranial nerve morbidity or at least prevent additional deficits compared with surgical resection. Twenty-two patients had previously undergone surgery and all had pretreatment cranial neuropathies. Median follow-up was 83 months, one of the longest in the radiosurgery literature for schwannoma.
Clinically, cranial nerve deficits improved in 20% and were stable in 77% after radiosurgery, but worsened in one patient. All nerves that were functioning pretreatment were intact after treatment (i.e., no new cranial nerve deficits). Radiographically, tumors regressed in 17 patients, were stable in 16, and progressed in 2. The authors submitted that their expectation, in terms of long-term tumor control and preservation or improvement of cranial nerve function, was confirmed.
The Current Role of Stereotactic Radiosurgery for Schwannomas Stereotactic radiosurgery clearly has a role in the treatment of nonvestibular skull base schwannoma, in both primary and adjuvant settings. The specifics of this role, however, remain controversial. It is possible to achieve high tumor control rates and significant symptom relief in many patients. The challenge is to identify who will benefit, who will not, and perhaps most importantly, who will be likely to be harmed in the process. With stereotactic radiosurgery, as with any modality of treatment for these challenging problems, complications can be devastating. In contrast with surgery, for irradiated patients the risk of serious complications extends well into the future. Late-onset brainstem edema, facial palsy and other cranial nerve deficits, cystic degeneration, and progressive fibrosis have all been reported. The risk of radiation-induced malignancy should be low with radiosurgery, but will not be accurately known for years to come. The incidence of radionecrosis affecting the brain and skull would also be expected to be low, but late-effects outcomes will take years if not decades to define. Currently, we have little means of identifying which patients will develop these long-term problems, but the concern is real, especially when radiation is considered for younger patients. With regard to radiation-related cranial nerve palsies, it is known that these phenomena are dose-dependent. For effective tumor control, treatment protocols most commonly employ a prescription dose, or so-called marginal tumor dose, in the range of 12 to 13 Gy (115) but significantly higher doses have been used in the past and continue to be used in some centers. Miller et al. (125) reported the Mayo Clinic experience with a reduced-dose protocol in treatment of vestibular schwannomas, in a study designed to lower the incidence of radiation-related cranial neuropathies. Comparing cohorts treated with standard doses vs. reduced doses, they found that the incidence of facial nerve morbidity and trigeminal nerve morbidity could be substantially reduced when the tumor-margin dose was lowered. This was statistically quite significant for facial palsy in particular, in which facial neuropathy went from 38% to 8%. During the median follow-up of 2.3 years, there were no cases of tumor progression in the reduced-dose group, but the authors noted that longer follow-up is needed to determine whether reduction of therapeutic dose will reduce ultimate tumor control rates. Also, as noted before, much is known about radiosurgery for VS, but much remains to be learned, and it is unclear to what degree VS data apply to NVS patients. A multitude of detailed issues beyond the scope of this discussion will continue to deserve scrutiny. As in Miller’s study, precise dose comparisons are needed to determine optimal “compromise dose” strategies that give good tumor control with minimal risk of complications. Comparison of other parameters, including tumor volumes, geometric tumor variations, and specific risks that vary by anatomic site must all be carefully considered.
Chapter 37: Schwannomas of the Skull Base
DECISION-MAKING IN THE MANAGEMENT OF SKULL BASE SCHWANNOMAS The rarity of nonvestibular skull base schwannomas does not diminish the grave implications of their diagnosis. These are devastating clinical problems. Perhaps it is this reality that has led to the vast number of publications on the subject, the number of which almost certainly exceeds the number of patients with the disease. Clearly, there is no perfect treatment that can be universally applied to all schwannoma patients. As outlined in this chapter, there are numerous factors that must be taken into account in order to determine the best course of action for a particular patient. The fact that a patient has schwannoma does not necessarily mandate treatment. Observation, or expectant management, may be the best option for select patients with skull base schwannoma. If observation is to be a serious consideration, it is essential that the diagnosis be established with as much certainty as the situation permits, using clinical, radiographic and (when appropriate) invasive means. Observation may be the best option for patients who are asymptomatic or minimally symptomatic; who are advanced in age or have medical contraindications to, or are unwilling to accept, active intervention; who would be likely to experience treatment-related deficits that are worse than their existing symptoms; or who have one of the neurofibromatosis syndromes in which multiple tumors exist and a new treatment-related deficit would be debilitating. Observation is not a passive choice. The decision to follow this course requires a firm commitment to detailed clinical and radiographic follow-up, at regular intervals and over the long term, so that tumor progression can be detected early enough to intervene with the least possible morbidity. As we have noted, there will be some patients who may never need intervention, but clearly many will. Intervention, be it with surgery or radiation, should be considered when the patient under observation becomes increasingly symptomatic, or when significant tumor progression is noted radiographically. One of the challenges lies in deciding what amount of growth is significant. Observation is not a risk-free choice. One potential problem with observation is the fact that some tumors will not follow the steady course of slow progression, and new deficits may come on quite suddenly. This may rarely be from a sudden tumor growth, or it may be due to intratumoral hemorrhage or rapid cyst formation. Deficits brought on under these circumstances may not be reversible. Surgical treatment of NVS historically carried with it a high incidence of morbidity, especially with respect to cranial nerve deficits. Substantial progress in the areas of microsurgical technique and cranial nerve rehabilitation has drastically reduced the morbidity of surgery when performed by expert skull base surgery teams. Many patients with complex tumor problems can be offered the possibility of complete tumor removal and improvement in cranial nerve function. Unquestionably, surgery is the first treatment of choice for very large or predominantly cystic schwannomas, or for those that compromise the brainstem or 4th ventricle. Even in cases where total resection is inadvisable or impossible, decompression or subtotal resection can greatly improve the patient’s outcome. A growing body of evidence suggests that stereotactic radiosurgery is a very reasonable and effective option for patients who have small NVS tumors that are noncystic and not in close proximity to the brainstem or 4th ventricle. It is both logical and appealing to assume that the generally high success rates that have been reported for vestibular schwannoma
535
will be applicable to the NVS population. The recent literature is encouraging in that regard, but experience is still limited. Given the rarity of NVS, multi-institutional prospective trials, or at least collaborative database sharing, are needed in order to further define the optimal use of this modality specifically for NVS. Currently, as a practical matter, treatment decisions should best be made in the context of a multidisciplinary discussion in which patients have the benefit of expert consultation from all relevant specialists. As Wackym stated, “An informed decision to pursue observation, microsurgery, stereotactic radiosurgery, or a combination of these methods must be made, and it remains the responsibility of the surgeon to provide a balanced view of the relative advantages and disadvantages of each method.” (126)
ACKNOWLEDGEMENT The senior author (DWN) gratefully acknowledges the following individuals for their roles in the genesis of this chapter: Wilma Rowley Nuss, Michael Artigues, Danny Douglass, Eric Dusang, Stacey Liljeberg, and Les Nettleton. REFERENCES 1. Batsakis JG. Tumors of the peripheral nervous system. In: Batsakis JG, ed. Tumors of the Head and Neck: Clinical and Pathological Considerations. Baltimore: Williams & Wilkins, Baltimore, 1982:313–333. 2. Katz AD, Passy V, Kaplan L. Neurogenous neoplasms of major nerves of face and neck. Arch Surg. 1971;103:51. 3. Kao G. Neurolemmoma. Nov. 27 , 2007. 4. Asthagiri AR, Helm GA, Sheehan JP. Current concepts in management of meningiomas and schwannomas. Neurol Clin North Am. 2007;25(4):1209–1230. 5. El-Mofty SK, Kyriakos M. Soft tissue and bone lesions. In: Gnepp DR, ed. Diagnostic Surgical Pathology of the Head and Neck. Philadelphia: W.B. Saunders, 2001:520–525. 6. MacCollin M, Woodfin W, Kronn D, et al. Schwannomatosis: A clinical and pathologic study. Neurology. 1997;48(3):787–788. 7. Mrugala MM, Batchelor TT, Plotkin SR. Peripheral and cranial nerve sheath tumors. Curr Opin Neurol. 2005;18(5):604–610. 8. Hanemann CO, Evans DG. News on the genetics, epidemiology, medical care and translational research of Schwannomas. J Neurol. 2006;253(12):1533–1541. 9. Fraenzer JT, Pan H, Minimo L Jr, et al. Overexpression of the NF2 gene inhibits schwannoma cell proliferation through promoting PDGFR degradation. Int J Oncol. 2003;23(6):1493–1500. 10. Evans DG, Moran A, King A, et al. Incidence of vestibular schwannoma and neurofibromatosis 2 in the North West of England over a 10-year period: Higher incidence than previously thought. Otol Neurotol. 2005;26:93–97. 11. Antinheimo J, Sankila R, Carpen O, et al. Population-based analysis of sporadic and Type 2 neurofibromatosis-associated meningiomas and schwannomas. Neurology. 2000;54:71–76. 12. MacCollin M, Chiocca EA, Evans DG, et al. Diagnostic criteria for schwannomatosis. Neurology. 2005;64(11):1838–1845. 13. Kahn AN. Neurofibromatosis: Conference statement. National Institutes of Health Consensus Development Conference. Arch Neurol. 1988;45(5):575–578. 14. Brooks DG. The neurofibromatoses: Hereditary predisposition to multiple peripheral nerve tumors. Neurosurg Clin N Am. 2004;15(2):145–155. 15. Ahn JY, Kwon SO, Shin MS, et al. A case of multiple schwannomas of the trigeminal nerves, acoustic nerves, lower cranial nerves, brachial plexuses and spinal canal: Schwannomatosis or neurofibromatosis? Yonsei Med J. 2002;43(1):109–113. 16. Purcell SM, Dixon SL. Schwannomatosis. An unusual variant of neurofibromatosis or a distinct clinical entity? Arch Dermatol. 1989;125(3):390–393.
536
Nuss and Burke
17. Huang JH, Simon SL, Nagpal S, et al. Management of patients with schwannomatosis: Report of six cases and review of the literature. Surg Neurol. 2004;62(4):353–361; discussion 361. 18. Kam J, Helm T. Neurofibromatosis. www.emedicine.com, January 2, 2007. 19. Skovronsky DM, Oberholtzer JC. Pathologic classification of peripheral nerve sheath tumors. Neurosurg Clin North Am. 2004;15:157–166. 20. Dodd LG, Marom EM, Dash RC, et al. Fine-needle aspiration cytology of "ancient" schwannoma. Diagn Cytopathol. 1999;20(5):307–311. 21. Maertens O, Brems H, Vandesompele J, et al. Comprehensive NF1 screening on cultured Schwann cells from neurofibromas. Hum Mutat. 2006;27(10):1030–1040. 22. Stout AP. Tumors of the peripheral nervous system. In: Atlas of Tumor Pathology, Section 2, Fascicle 6. Armed Forces Institute of Pathology, Washington, D.C., 1946. 23. Preston FW, Walsh WS, Clarke TH. Cutaneous neurofibromatosis: Clinical manifestations and incidence of sarcoma in 61 male patients. Arch Surg. 1952;64:813. 24. Gupta G, Maniker A. Malignant peripheral nerve sheath tumors. Neurosurg Focus. 200715;22(6):E12. 25. Ferrari A, Bisogno G, Carli M. Management of childhood malignant peripheral nerve sheath tumor. Paediatr Drugs. 2007;9(4):239–248. 26. Minovi A, Basten O, Hunter B, et al. Malignant peripheral nerve sheath tumors of the head and neck: Management of 10 cases and literature review. Head Neck. 2007;29(5):439–445. 27. Huang SQ, Liang BL, Yuan JP, et al. MRI performance and diagnosis of schwannomas from cranial nerves in posterior cranial fossae. Ai Zheng. 2006;25(9):1178–1182. 28. Eldevik OP, Gabrielsen TO, Jacobsen EA. Imaging findings in schwannomas of the jugular foramen. Am J Neuroradiol. 2000;21(6):1139–1144. 29. Fisher LM, Doherty JK, Lev MH, et al. Distribution of nonvestibular cranial nerve schwannomas in neurofibromatosis 2. Otol Neurotol. 2007;28(8):1083–1090. 30. Mautner VF, Lindenau M, Baser ME, et al. The neuroimaging and clinical spectrum of neurofibromatosis 2. Neurosurgery. 1996;38(5):880–885; discussion 885–886. 31. Nunes F, MacCollin M. Neurofibromatosis 2 in the pediatric population. J Child Neurol. 2003;18(10):718–724. 32. Kouyialis AT, Stranjalis G, Papadogiorgakis N, et al. Giant dumbbell-shaped middle cranial fossa trigeminal schwannoma with extension to the infratemporal and posterior fossae. Acta Neurochir. 2007;149(9):959–964. 33. Eisenman D, Voigt E, Selesnick S. Unusual tumors of the internal auditory canal and cerebellopontine angle. In: Jacker R, Driscoll CLW, eds. Tumors of the Ear and Temporal Bone. Philadelphia: Lippincott Williams & Wilkins, 2000:259. 34. Nager GT. Neurinomas of the trigeminal nerve. Am J Otolaryngol. 1984;5:301–333. 35. Samii M, Migliori MM, Tatagiba M, et al. Surgical treatment of trigeminal schwannomas. J Neurosurg. 1995;82(5):711–718. 36. Pamir MN, Peker S, Bayrakli F, et al. Surgical treatment of trigeminal schwannomas. Neurosurg Rev. 2007;30(4):329–337; discussion 337. epub 2007, Aug 4. 37. Gwak HS, Hwang SK, Paek SH, et al. Long-term outcome of trigeminal neurinomas with modified classification focusing on petrous erosion. Surg Neurol. 2003;60(1):39–48; discussion 48. 38. Verstappen CC, Beems T, Erasmus CE, et al. Dumbbell trigeminal schwannoma in a child: Complete removal by a one-stage pterional surgical approach. Childs Nerv Syst. 2005;21(11):1008– 1011. Epub 2005 Mar 16. 39. Zhou LF, Mao Y, Zhang R. Surgical treatment of dumbbellshaped neurinomas: Report of an experience with 57 cases in a single hospital. Surg Neurol. 2007;68(6):594–602. 40. Al-Mefty O, Ayoubi S, Gaber E. Trigeminal schwannomas: Removal of dumbbell-shaped tumors through the expanded Meckel cave and outcomes of cranial nerve function. J Neurosurg. 2002;96(3):453–463. 41. Boustred AM. Endoscopic excision of a supraorbital schwannoma. Ophthal Plast Reconstr Surg. 2007;23(1):66–68.
42. Roh JL. Removal of infratemporal fossa schwannoma via a transmandibular transpterygoid approach. Eur Arch Otorhinolaryngol. 2005;262(5):428–431. Epub 2004 Sep 18. 43. Park EH, Lee SS, Byun SW. A schwannoma in the nasal septum. Eur Arch Otorhinolaryngol. 2007 [Epub ahead of print]. 44. Sciarretta V, Pasquini E, Frank G, et al. Endoscopic treatment of benign tumors of the nose and paranasal sinuses: A report of 33 cases. Am J Rhinol. 2006;20(1):64–71. 45. Har-El G. Combined endoscopic transmaxillary-transnasal approach to the pterygoid region, lateral sphenoid sinus, and retrobulbar orbit. Ann Otol Rhinol Laryngol. 2005;114(6):439– 442. 46. Kennerdell JS, Maroon JC, Celin SE. The posterior inferior orbitotomy. Ophthalmic Plastic and Reconstr Surg. 1998;14(4):277–280. 47. Eisenberg MB, Al-Mefty O, DeMonte F, et al. Benign nonmeningeal tumors of the cavernous sinus. Neurosurgery. 1999;44(5):949–954; discussion 954–955. 48. Cockerham KP, Cockerham GC, Stutzman R, et al. The clinical spectrum of schwannomas presenting with visual dysfunction: A clinicopathologic study of three cases. Surv Ophthalmol. 1999;44(3):226–234. 49. Cantore WA. Neural orbital tumors. Curr Opin Ophthalmol. 2000;11(5):367–371. 50. Moffat D, De R, Hardy D, et al. Surgical management of trigeminal neuromas: A report of eight cases. J Laryngol Otol. 2006;120(8):631–637. Epub 2006 May 17. 51. Hajjaj M, Linthicum FH Jr. Facial nerve schwannoma: Nerve fibre dissemination. J Laryngol Otol. 1996;110(7):632–633. 52. Dort JC, Fisch U. Facial nerve schwannomas skull base surg. 1991;51–56. 53. Schaitkin BM. Facial nerve schwannoma. In: Jackler R, Driscoll CLW, eds. Tumors of the Ear and Temporal Bone. Philadelphia: Lippincott Williams & Wilkins, 2000: 276–289. 54. O’Donaghue GM, Brackmann DE, House JW, et al. Neuromas of the facial nerve. Am J Otol. 1989;10:49–54. 55. Wiggins RH III, Harnsberger HR, Salzman KL, et al. The many faces of facial nerve schwannoma. AJNR Am J Neuroradiol 2006;27:694–699. 56. Kreeft A, Schellekens PP, Leverstein H. Intraparotid facial nerve schwannoma. What to do? Clin Otolaryngol. 2007;32(2):125– 129. 57. Perez R, Chen JM, Nedzelski JM. Intratemporal facial nerve schwannoma: A management dilemma. Otol Neurotol. 2005;26(1):121–126. (Toronto) 58. Elahi M, Audet N, Rochon L. Intraparotid facial nerve schwannoma. J Otolaryngol. 1995;24:364–367. 59. Lipkin AF, Coker NJ, Jenkins HA, et al. Migraine and sudden sensorineural hearing loss. Arch Otolaryngol Head Neck Surg. 1987;113:325–326. 60. Fisch V, Ruttner J. In: Fisch U, ed. Pathology of Intratemporal Tumors Involving the Facial Nerve. The Netherlands: Kugler Medical Publishing, 1997. 61. McMenomey SO, Glasscock ME 3rd, Minor LB, et al. Facial nerve neuromas presenting as acoustic tumors. Am J Otol. 1994;15(3):307–312. 62. Park HY, Kim SH, Son EJ, et al. Intracanalicular facial nerve schwannoma. Otol Neurotol. 2007;28(3):376–380. 63. Kubota Y, Kawamata T, Kubo O, et al. Large facial nerve schwannomas without facial palsy: Case reports and review of the literature. Neurosurg Rev. 2005;28(3):234–238. Epub 2005 Mar. 64. Shimizu K, Iwai H, Ikeda K, et al. Intraparotid facial nerve schwannoma: A report of five cases and an analysis of MR imaging results. AJNR Am J Neuroradiol. 2005;26(6):1328–1330. 65. Caughey RJ, May M, Schaitkin BM. Intraparotid facial nerve schwannoma: Diagnosis and management. Otolaryngol Head Neck Surg. 2004;130(5):586–592. 66. Kertesz TR, Shelton C, Wiggins R, et al. Intratemporal facial nerve neuroma: Anatomical location and radiological features laryngoscope. 2001;111(7):1250–1256. 67. Litre CF, Gourg GP, Tamura M, et al. Gamma knife surgery for facial nerve schwannomas. Neurosurgery. 2007;60(5):853–859; discussion 853–859.
Chapter 37: Schwannomas of the Skull Base 68. Guermazi A, Lafitte F, Miaux Y, et al. The dural tail sign: Beyond meningioma. Clin Radiol. 2005; 60:171–188. 69. Stasolla A, Bellussi A, Bibbolino C. Dural tail: Another face of facial nerve schwannoma? AJNR Am J Neuroradiol. 2006;27(3):694–699. 70. Sataloff RT, Frattali MA, Myers DL. Intracranial facial neuromas: Total tumor removal with facial nerve preservation: A new surgical technique. Ear Nose Throat J. 1995;74:244–246, 248–256. 71. Liu R, Fagan P. Facial nerve schwannoma: Surgical excision versus conservative management. Ann Otol Rhinol Laryngol. 2001;110(11):1025–1029. 72. Valentino J, Boggess MA, Ellis JL, et al. Expected neurologic outcomes for surgical treatment of cervical neurilemomas. Laryngoscope. 1998;108:1009–1013. 73. Kim CS, Chang SO, Oh SH, et al. Management of intratemporal facial nerve schwannoma. Otol Neurotol. 2003;24(2):312–316. 74. Conley J, Janecka I. Schwann cell tumors of the facial nerve. Laryngoscope. 1974;84(6):958–962. 75. Lee JD, Kim SH, Song MH, et al. Management of facial nerve schwannoma in patients with favorable facial function. Laryngoscope. 2007;117(6):1063–1068. 76. Angeli SI, Brackmann DE. Is surgical excision of facial nerve schwannomas always indicated? Otolaryngol Head Neck Surg 1997;117:S144–S147. 77. Von Doersten PG. Jugular foramen schwannoma. In: Jacker R, Driscoll CLW, eds. Tumors of the Ear and Temporal Bone. Philadelphia: Lippincott Williams & Wilkins, 2000:374–387. 78. Tatebayashi K, Tanaka Y, Numata H, et al. Schwannoma of the spinal accessory nerve in the cisterna magna. Surg Neurol. 2003;59(3):217–222; discussion 222. 79. Boulton MR, Cusimano MD. Foramen magnum meningiomas: Concepts, classifications, and nuances. Neurosurg Focus. 2003;14(6): Article 10. 80. Kurokawa R, Tabuse M, Yoshida K, et al. Spinal accessory schwannoma mimicking a tumor of the fourth ventricle: Case report. Neurosurgery. 2004;54(2):510–514; discussion 514. 81. Kaye AH, Hahn JF, Kinney SE, et al. Jugular foramen schwannomas. J Neurosurg. 1984;60:1045–1053. 82. Pellet W, Cannoni M, Perch A. The widened transcochlear approachto jugular foramen tumors. J Neurosurg. 1988;69:887– 894. 83. Franklin DJ, Moore GF, Fisch U. Jugular foramen peripheral nerve sheath tumors. Laryngoscope. 1989;99:1081–1087. 84. House JW, Brackmann DE. Facial nerve grading system. Otolaryngol Head Neck Surg. 1985;93:146–147. 85. Wilson MA, Hillman TA, Wiggins RH, et al. Jugular foramen schwannomas: Diagnosis, management, and outcomes. Laryngoscope. 2005;115(8):1486–1492. 86. Sanna M, Bacciu A, Falcioni M, et al. Surgical management of jugular foramen schwannomas with hearing and facial nerve function preservation: A series of 23 cases and review of the literature. Laryngoscope. 2006;116(12):2191–2204. 87. Kadri PAS, Al-Mefty O. Surgical treatment of dumbbellshaped jugular foramen schwannomas. Neurosurgical Focus. 2004;17(2):56–62. 88. Ahmad FU, Gupta A, Sharma MC, et al. The enigmatic origin of subfrontal schwannomas: Report of a case without hyposmia. Acta Neurochir (Wien). 2006;148(6):671–672; discussion 672. Epub 2006 Feb 9. 89. Adachi K, Yoshida K, Miwa T, et al. Olfactory schwannoma. Acta Neurochir. 2007;149(6):605–611. 90. Kinoshita Y, Kurosaki M, Kamitani H, et al. Schwannoma originating in the optic canal. Acta Neurochir (Wien). 2008;150(1):89– 90; discussion 90. Epub 2008 Jan 10. 91. Kim DS, Choi JU, Yang KH, et al. Optic sheath schwannomas: Report of two cases. Childs Nerv Syst. 2002;18(12):684–689. Epub 2002 Oct 16. 92. Wewetzer K, Verdu´ E, Angelov DN, et al. Olfactory ensheathing glia and Schwann cells: Two of a kind? Cell Tissue Res. 2002;309(3):337–345. Epub 2002 Jul 17. 93. Thompson RJ, Roberts B, Alexander CL, et al. Comparison of neuregulin-1 expression in olfactory ensheathing cells, Schwann cells and astrocytes. J Neurosci Res. 2000;61(2):172–185.
537
94. Yuen A, Trost N, McKelvie P, et al. Subfrontal schwannoma: A case report and literature review. J Clin Neurosci. 2004;11(6):663–665. 95. Miller NR. Primary tumours of the optic nerve and its sheath. Eye. 2004;18(11):1026–1037. 96. Tanriover N, Kemerdere R, Kafadar AM, et al. Oculomotor nerve schwannoma located in the oculomotor cistern. Surg Neurol. 2007;67(1):83–88; discussion 88. Epub 2006 Nov 3. 97. Ohata K, Takami T, Goto T, et al. Schwannoma of the oculomotor nerve. Neurol India. 2006;54(4):437–439. 98. Mariniello G, Horvat A, Dolenc VV. En bloc resection of an intracavernous oculomotor nerve schwannoma and grafting of the oculomotor nerve with sural nerve. Case report and review of the literature. J Neurosurg. 1999;91(6):1045–1049. Review. 99. Gerganov V, Amir S, Koerbel A, et al. Cystic trochlear nerve schwannoma. Case report. Surg Neurol. 2007;68(2):221–225. Epub 2007 Jun 21. 100. Ohba S, Miwa T, Kawase T. Trochlear nerve schwannoma with intratumoral hemorrhage: Case report. Neurosurgery. 2006;58(4):E791; discussion E791. 101. Nakamura M, Carvalho GA, Samii M. Abducens nerve schwannoma: A case report and review of the literature. Surg Neurol. 2002;57(3):183–188; discussion 188–189. 102. Acharya R, Husain S, Chhabra SS, et al. Sixth nerve schwannoma: A case report with literature review. Neurol Sci. 2003;24(2):74–79. 103. Nakagawa T, Uchida K, Ozveren MF, et al. Abducens schwannoma inside the cavernous sinus proper: Case report. Surg Neurol. 2004;61(6):559–563; discussion 563. 104. Ho CL, Deruytter MJ. Navigated dorsolateral suboccipital transcondylar (NADOSTA) approach for treatment of hypoglossal schwannoma. Case report and review of the literature. Clin Neurol Neurosurg. 2005;107(3):236–242. 105. Sato K, Shimizu S, Oka H, et al. Usefulness of transcervical approach for surgical treatment of hypoglossal schwannoma with paraspinal extension: Case report. Surg Neurol. 2006;65(4):397– 401, discussion 401. 106. Tucker A, Miyake H, Tsuji M, et al. Intradural microsurgery and extradural gamma knife surgery for hypoglossal schwannoma: Case report and review of the literature. Minim Invasive Neurosurg. 2007;50(6):374–378. 107. Halefoglu ˘ AM. Neurofibromatosis type 2 associated with multiple cranial nerve schwannomas: A case report. Kulak Burun Bogaz Ihtis Derg. 2007;17(3):171–175. 108. Cheong JH, Kim JM, Bak KH, et al. Bilateral vidian nerve schwannomas associated with facial palsy. Case report and review of the literature. J Neurosurg. 2006;104(5):835–839. 109. Wittekindt C, Liu WC, Hampl JA, et al. Radiology quiz case 1: Intracranial schwannoma originating from the greater petrosal nerve. Arch Otolaryngol Head Neck Surg. 2006;132(1):102,104. 110. Vasama JP, Rahnasto J, Dastidar P. Jacobson’s nerve schwannoma. Otol Neurotol. 2007;28(1):141. ¨ U, Seker A, Kurtkaya O, et al. Internal carotid plexus 111. Ture schwannoma of the cavernous sinus: Case report. Neurosurgery. 2003;52(2):435–438; discussion 438–439. 112. Kamel MH, Mansour NH, Mascott C, et al. Compression of the rostral ventrolateral medulla by a vagal schwannoma of the cerebellomedullary cistern presenting with refractory neurogenic hypertension: Case report. Neurosurgery. 2006;58(6):E1212; discussion E1212. 113. Jagetia A, Pratap Kumar N, Singh D, et al. Pathological laughter in trigeminal schwannoma: Case report and review of the literature. Neurosurg Rev. 2006;29(4):348–352; discussion 352. Epub 2006 Sep 5. 114. Hsu WM, Wang AG. Pneumosinus dilatans associated with orbital neurilemmoma. Zhonghua Yi Xue Za Zhi (Taipei). 2002;65(5):218–224. 115. Flickinger JC, Barker FG 2nd. Clinical results: Radiosurgery and radiotherapy of cranial nerve schwannomas. Neurosurg Clin N Am. 2006;17(2):121–128, vi. 116. Pollock BE, Foote RL, Stafford SL. Stereotactic radiosurgery: The preferred management for patients with nonvestibular schwannomas? Int J Radiat Oncol Biol Phys. 2002;52(4):1002–1007.
538
Nuss and Burke
117. Sheehan J, Yen CP, Arkha Y, et al. Gamma knife surgery for trigeminal schwannoma. J Neurosurg. 2007;106 (5):839– 845. 118. Hasegawa T, Kida Y, Yoshimoto M, et al. Trigeminal schwannomas: Results of gamma knife surgery in 37 cases. J Neurosurg. 2007;106(1):18–23. 119. Huang CF, Kondziolka D, Flickinger JC, et al. Stereotactic radiosurgery for trigeminal schwannomas. Neurosurgery. 1999;45(1):11–16; discussion 16. 120. Pan L, Wang EM, Zhang N, et al. Long-term results of Leksell gamma knife surgery for trigeminal schwannomas. J Neurosurg. 2005;102 (Suppl):220–224. 121. Peker S, Bayrakli F, Kilic¸ T, et al. Gamma-knife radiosurgery in the treatment of trigeminal schwannomas. 2007;149(11):1133– 1137; discussion 1137. Epub 2007, Aug 31.
122. Akiyama T, Ikeda E, Kawase T, et al. Pseudocapsule formation after gamma knife radiosurgery for trigeminal neurinoma— case report. Neurol Med Chir (Tokyo). 2005;45(10):526–529. 123. Kida Y, Yoshimoto M, Hasegawa T. Radiosurgery for facial schwannoma. J Neurosurg. 2007;106(1):24–29. 124. Martin JJ, Kondziolka D, Flickinger JC, et al. Cranial nerve preservation and outcomes after stereotactic radiosurgery for jugular foramen schwannomas. Neurosurgery. 2007; 61(1): 76– 81. 125. Miller RC, Foote RL, Coffey RJ, et al. Decrease in cranial nerve complications after radiosurgery for acoustic neuromas: A prospective study of dose and volume. Int J Radiat Oncol Biol Phys. 1999;43(2):305–311. 126. Wackym PA. Stereotactic radiosurgery, microsurgery, and expectant management of acoustic neuroma: Basis for informed consent. Otolaryngol Clin North Am. 2005;38(4):653–670.
38 Paragangliomas of the Head and Neck David P. Goldstein, Mark G. Shrime, Bernard Cummings, and Patrick J. Gullane
(nodose) ganglion (5). They may be found within or below the perineurium of the vagus nerve, or dispersed among the nerve fibers (4,5). Smaller collections of paraganglia are located within the larynx, in the supraglottis (superior paraganglia), and in the subglottis, in close proximity to the cricoid (inferior paraganglia) (16).
INTRODUCTION Paragangliomas of the head and neck represent a rare group of neoplasms, arising from paraganglionic tissue located throughout the head and neck. They comprise 0.6% of all head and neck tumors, and only 0.01% to 0.03% of all tumors diagnosed in humans (1,2). Incidence in the general population ranges from 1 in 30,000 to 1 in 100,000 (1,3). Only 3% of all paragangliomas arise in the head and neck (4,5). Since the early 20th century there have been numerous reports of paragangliomas arising at various locations throughout the head and neck. Von Haller’s description of the carotid body in 1743 marks the first recorded mention of paraganglionic tissue in the literature (6). In 1862, von Luschka described tumors arising from the carotid body; surgical excision of which was first reported by Scudder in 1903 (6,7). In 1946, Rossenwasser described removal of a “carotid body tumor” of the middle ear and proposed that the tumor may have arisen from recently described glomus bodies in the temporal bone (8,9). The first description of a vagal paraganglioma was by Stout in 1935 (10).
Physiology and Function of Paraganglia Paraganglia contain two cell types: chief cells and sustentacular cells, which organize in clusters called zellballen (17). Chief cells are filled with catecholamines and tryptophanrich proteins (17). Within the chief cells of the extra-adrenal paraganglia, only norepinephrine is present, because these cells lack methyltransferase, the enzyme necessary for the conversion of norepinephrine to epinephrine (18). Sustentacular cells function as support cells. The rich microvasculature within paraganglia facilitates secretion of the granular products into the bloodstream (17). During embryogenesis, paraganglia serve as the major source of catecholamines (19). In adulthood, their role changes, and they become chemoreceptors, responding to alterations in homeostasis (20,21). The carotid body is involved in the reflex regulation of arterial pH, pO2 and pCO2 . The role of the other paraganglia is not well understood, but it may be a similar role to that of the carotid body (11).
THE PARAGANGLION SYSTEM Paraganglia, aggregates of cells located within neuronal and vascular adventitia throughout the body (11), are part of the diffuse neuroendocrine system, or amine precursor decarboxylate system (12). These cells arise from the neural crest and migrate during embryogenesis to concentrate around autonomic ganglia (3). The largest collection remains in the adrenal medulla (3), which secretes catecholamines in association with the sympathetic nervous system (11). Paraganglia located within the head and neck are associated with the parasympathetic nervous system (11). The carotid bodies constitutes the largest collection of paraganglia in the head and neck and are located in the adventia or periadvential tissue of the posteromedial wall of the carotid bifurcation (6). In the vast majority of patients, a single carotid body is found at each carotid bifurcation (13,14). In the temporal bone paraganglia are found arising along Arnold nerve (the auricular branch of the vagus nerve), and Jacobson nerve (the tympanic branch of the glossopharyngeal nerve), and juxtaposed to the jugular bulb (15). On average, there are three such glomus bodies in each ear, although the number decreases after the age of 60 (15). Slightly more than 50% of temporal bone paraganglia are located in the region of the jugular fossa, approximately 30% are within the mucosa of the cochlear promontory, and 10% lie within the inferior tympanic canaliculus (15). Vagal paraganglia may occur anywhere along the vagus nerve but are most commonly located around the inferior
Paraganglioma Nomenclature Paragangliomas are benign neuroendocrine tumors arising from the paraganglionic system (22). These tumors have, at various times, been called nonchromaffin tumors, chemodectomas, glomus tumors, and carotid body tumors. They have been described by their morphology (glomus refers to the tuft-like appearance of the tumor vasculature), by the physiologic function of the organ from which they derive (chemodectoma is derived from the Greek ch¯emeia- “chemical” and dektos- “to receive”), or by their histological properties (e.g., chromaffin vs. nonchromaffin) (11). Glenner and Grimley developed a classification system based on embryology, anatomic location, and histology which is presented in Table 1 (4,18). Under the World Health Organization Classification of Tumors, paragangliomas are classified by their anatomic location (Table 2) (11,23). Currently, the latter is the preferred classification system.
Location and Routes of Spread Paragangliomas have been described in nearly 20 distinct locations in the head and neck, with the carotid body being the most common location (12). Carotid body and jugulotympanic paragangliomas account for 80% of all head and neck paragangliomas (24) and vagal paragangliomas account for another 5% (1,5,25). Less common head and neck sites include 539
540
Goldstein et al. Table 1 Glenner and Grimley Classification Adrenal
Extra-adrenal
Pheochromocytoma
Branchiomeric Aorticopulmonary Coronary Intercarotid Jugulotympanic Laryngeal Nasal Orbital Pulmonary Subclavian Intravagal Aorticosympathetic Visceroautonomic
the trachea (26), larynx (27–29), paranasal sinuses (30–32), orbit (33), sympathetic trunk (34), and thyroid (35). Carotid body tumors (CBTs), as their nomenclature implies, arise from the paraganglia forming the carotid body (6,36). Their blood supply is derived principally from branches of the external carotid artery (ECA); they can, however, receive blood supply from branches of other major vessels, including the internal carotid (ICA) and the vertebral arteries (37). These tumors typically cause splaying of the carotid bifurcation, displacing the ICA posterolaterally and the ECA anterolaterally or anteromedially. As they enlarge, they tend to encase the ICA and ECA, without narrowing their caliber (5). Based on their location, growth may involve surrounding nerves such as the hypoglossal and vagus nerves or the sympathetic chain, or may impinge on the skull base itself. Medial extension into the parapharyngeal space occurs in up to 20% of cases (38,39). Tympanic paragangliomas develop along the tympanic canaliculus, in association with Jacobson nerve, and over the cochlear promontory within the tympanic cavity of the middle ear, in association with Arnold nerve. As these tumors enlarge, they surround the ossicles, fill the tympanic cavity, occlude the eustachian tube, and expand through the aditus ad antrum into the mastoid cavity. They may also protrude through the tympanic membrane into the external auditory canal (15). Large tumors can extend inferiorly to involve the jugular foramen, making them difficult to differentiate from jugular paragangliomas. Although bone destruction is unusual (40–42), these tumors can involve the cochlea, facial nerve, jugular bulb, sigmoid sinus, and petrous carotid (43). Tympanic paragangliomas typically derive their vascular supply from the inferior tympanic artery. Arterial supply may also come from other branches of the ICA and ECA supplying the middle ear and temporal bone (15). Jugular paragangliomas originate within the adventitia of the jugular bulb, in the lateral portion of the jugular foramen. As they grow, they may extend into the medial portion of the jugular canal. In the early stage of tumor growth, Table 2 WHO Classification of Extra-adrenal Paraganglioma of the Head and Neck Carotid body Jugulotympanic Vagal Laryngeal Aortico-pulmonary Orbital nasopharyngeal
there may only be slight erosion of the bony cortex, usually along the lateral border of the jugular fossa (15). Continued growth leads to further destruction and irregular enlargement of fossa. Tumors may grow to involve the hypoglossal canal and nerve or the intrapetrous segment of the ICA (37). Intracranial extension (ICE) is postulated to occur via spread through the jugular or hypoglossal canals into the posterior fossa or through the carotid canal into the middle cranial fossa (44). The presence of ICE is reported to be between 14% and 72% (45,46). These tumors receive their blood supply primarily from the ascending pharyngeal artery but may also receive contributions from the occipital and post-auricular arteries, and, less frequently, branches of the vertebral and posterior inferior cerebellar (PICA) arteries (37). Vagal paragangliomas (VP) can arise at any point along the course of the vagus nerve from either the superior (jugular) ganglion within the jugular fossa, or the inferior (nodose) ganglion (15). Most commonly, they originate from the latter, approximately 2 cm below the jugular foramen (47). In contrast with CBTs, these tumors typically cause displacement of the ECA and ICA anteromedially, without splaying at the carotid bifurcation. Tumors of the inferior ganglion tend to grow to involve the post-styloid parapharyngeal space (12); superior growth may lead to involvement of the skull base and jugular foramen. Tumors arising in the middle or superior vagal ganglia, in contrast, are associated with early skull base involvement and intracranial extension (12). Further extension through the jugular foramen allows posterior fossa involvement (37). The vascular supply of vagal paragangliomas is derived from the occipital and ascending pharyngeal arteries. Laryngeal paragangliomas are very rare and most commonly arise from the superior paraganglia, located within the supraglottis (16,27). Paragangliomas of the inferior laryngeal paraganglion, depending on their anatomic location, may give rise to one of two different clinical entities, namely the so-called thyroid paragangliomas, which arise within the thyroid parenchyma itself (35), and subglottic paragangliomas which are in intimate association with the cricoid cartilage (48). Paragangliomas of the sinonasal cavity are also extremely rare (16). They have been reported in all sites within the nose and paranasal sinuses, including the frontal and sphenoid sinuses, and at the terminal portion of the pterygopalatine fossa, in close association with the ptyergoid ganglion (49).
Epidemiology and Etiology Head and neck paragangliomas tend to occur in women three to four times more frequently than in men (15,18,50,51), which is in contrast to paragangliomas of other extra-adrenal sites, in which men are more commonly affected (50). Two-thirds of patients are in the fourth and fifth decades of life when diagnosed, although age at diagnosis ranges from 6 months to 88 years (15,22). An increased incidence of CBTs have been noted in patients living with chronic sustained hypoxemia (52–56), most commonly in those living at high altitudes (22,54–56). Paragangliomas within this specific group of patients have an evident female predominance (8.3:1), low rate of bilaterality (5%), and a family history of 1% (55). Emerging evidence from genetic studies implicates heredity in 35% to 55% of individuals presenting with a head and neck paraganglioma (23,57). Prior estimates of 10% are likely miscalculations, related to the mode of genetic transmission. Paragangliomas are now recognized to be transmitted in a manner influenced by maternal imprinting, which
Chapter 38: Paragangliomas of the Head and Neck
can cause the phenotypic expression of germline mutations to skip multiple generations (57). Patients with familial paragangliomas have a significantly lower age of onset (25,57–62), a higher rate of multicentricity (57,58), and are more likely to have a CBT (59), or VP (63), than paragangliomas at other sites. Similarly, patients with CBTs are 5.8 times more likely to have familial tumors than those diagnosed with paragangliomas at other sites (59). Multicentricity occurs in 10% to 15% of nonfamilial cases (3,64); in familial cases, the incidence of multicentricity ranges from 25% to 87% of patients (12,25,62,65–68).
Genetics of Paragangliomas Familial paragangliomas may occur as part of a familial tumor syndrome or as isolated hereditary tumors (57). Familial syndromes that are known to be associated with the development of paragangliomas include von-Hippel-Lindau, Multiple Endocrine Neoplasia IIA and IIB , and the Carney triad (paragangliomas, pulmonary chondromas, and gastrointestinal stroma tumors) (19,23,25,69–75). Nonsyndromic familial paragangliomas are inherited in an autosomal dominant fashion, with genomic imprinting of the maternal allele (5,19,51,62,76). Genomic imprinting implies that affected men, who pass on the unimprinted genes, have a 50% chance of having an affected child, whereas affected women will not have affected children but can pass an inactivated gene to the next generation. A male child who inherits that gene will then produce children with a 50% chance of developing the tumor (51). Inheritance is currently believed to be related to mutations in the succinate dehydrogenase (SDH) gene (77). SDH is a mitochondrial enzyme complex, playing a role in key functions of the Krebs cycle, oxidative phosphorylation, and intracellular oxygen sensing and signaling (3,77– 79). Three of the four genes encoding subunits of the mitochondrial II complex—SDHB (pgl4 on 1p35-36), SDHC (pgl3 on 1q21), and SDHD (pgl1 on 11q23)—have been implicated in the pathogenesis of hereditary head and neck paragangliomas (63,77,80–82). It is thought that these mutations lead to a chronic hypoxic signal within the cell, causing cellular proliferation and tumor formation, a theory supported by the higher incidence of paragangliomas in patients living with chronic hypoxia (57,78,83). Schiavi et al. reviewed the prevalence of different mutations in 121 symptomatic, unrelated cases, in the International Head and Neck Paraganglioma Registry (84). The prevalence of SDHC mutation was 4%, SDHB was 7%, and SDHD was 17%. In three other non-population–based studies overall mutation frequencies ranged from 12% to 41% (63,81,82). In familial cases SDHD mutations account for 50% of cases, and SDHB mutations account for 20% (57,82). SDHD mutations have also been shown to predispose to the development of multifocal paragangliomas (85), while patients with mutations in SDHB are at increased risk for malignant paragangliomas (43,85–87) and may also be at increased risk for renal cell carcinoma and papillary thyroid carcinoma (85). Familial paragangliomas constitute approximately 20% of lesions for which genetic defects are known (3). Sporadic mutations in SDHB and SDHD each occur in less than 10% of all nonfamilial cases of paragangliomas (3,57,82,88), while the pathogenesis of the remaining 80% remains unknown (3).
Genetic Counseling Genetic counseling and/or radiologic screening should be considered in family members of patients with familial paragangliomas (57,58,72,89,90). It can help identify those at risk
541
of disease, allowing for early identification and treatment (57). Carriers could also be identified, allowing their offspring to benefit from genetic counseling (57). Dundee et al. recommend that genetic counseling be offered to all patients from the age of 5 who have a family history of paraganglioma (57). Those positive for the paternal PGL gene should then undergo radiological screening every 3 years. Family members of those with the carrier state should also be offered genetic counseling from 5 years of age. In those who choose not to undergo genetic counseling, radiologic screening should be offered from the age of 10 years. Patients presenting with sporadic disease should be offered genetic counseling, particularly those presenting at a young age or with multiple tumors. Genetic screening should then be extended to family members at risk of transmission.
Pathology Paragangliomas are solid neoplasms with a homogenous tan to red-brown appearance and may be partially or completely encapsulated (22,23). On microscopic examination, paragangliomas are composed predominantly of chief cells and sustentacular cells. Chief cells are arranged in distinctive clusters of cells, referred to as zellballen, and surrounded by extensive vascular sinusoids, sustentacular cells, and a stroma composed of a prominent fibrovascular tissue (Fig. 1) (2). Chief cells are round or oval cells with uniform nuclei, dispersed chromatin, and abundant, eosinophilic, granular or vacuolated cytoplasm. Sustentacular cells are spindle-shaped, basophilic cells (modified Schwann cells). Head and neck paragangliomas tend to have the same histologic appearance irrespective of their location. In most instances, tumor cells are relatively homogenous in their appearance. Infrequently, nuclear pleomorphism, necrosis, and increased mitotic activity may be found. These cellular features do not imply malignancy but may indicate a more aggressive neoplasm (51). On immunohistochemistry tumor cells are agyrophilic and cell nests may be delineated by reticular staining (51). The chief cells stain for chromogranin, synaptophysin, and neuron-specific enolase, while sustentacular cells stain positive for S-100 (22). Argentaffin, mucin, and periodic acid Schiff stains are negative (22). On electron microscopic examination, the hallmark of these tumors is presence of neurosecretory granules within the chief cells. The differential diagnosis of these tumors on light microscopy includes carcinoid tumors, neuroendocrine carcinomas, medullary thyroid carcinomas, middle ear adenomas, meningiomas, hemangiopericytomas, alveolar soft part sarcomas, and metastatic renal cell carcinomas (22).
Malignant Paragangliomas The diagnosis of malignant paragangliomas rests on clinical behavior rather than histologic appearance. Prominent sustentacular cells, necrosis, mitotic activity, nuclear pleomorphism, and perineural, bony, and vascular invasion may be seen in both benign and malignant paragangliomas (2,25,91,92). A paraganglioma is determined to be malignant only if metastases to non-neuroendrocine tissue can be demonstrated (Fig. 2) (2,25,91,92). Nodal metastases occur more commonly than distant metastases to the lungs, liver, bone, and skin (2,23,93,94). Examples of metastasizing paragangliomas have been reported in all head and neck locations from which paragangliomas arise (91). In general, less than 5% of all paragangliomas are malignant (2,91); however, prevalence depends on the site of the primary (12). Vagal paragangliomas appear to be associated with the highest rates
542
Goldstein et al.
Figure 1 Carotid body paraganglioma showing nests of clear cells surrounded by thin capillaries. This appearance represents the classical “Zellballen” pattern (A). The tumor cells are diffusely positive for the neuroendocrine marker synaptophysin (B). S-100 immunohistochemical stain highlighting peripheral sustentacular cells (C).
(10–19%) of malignancy among the more common head and neck sites (12,18,91,92), while orbital and laryngeal paragangliomas have slightly higher reported rates (20–25%) (12,95). Malignant CBTs and jugulotympanic paragangliomas have been reported to occur in about 3% to 6% of cases (7,91,96). In a National Cancer Database review of 59 cases of malignant paragangliomas, 68.6% were confined to regional nodes, and 31.4% of patients had distant disease (2). The overall 5-year relative survival rate was 59.5%. The 5-year survival rate for patients with metastases limited to lymph nodes was 76.8%, significantly higher than that for patients with distant metastases (11.8%) (2)
Secreting Paragangliomas Fewer than 5% of head and neck paragangliomas secrete catecholamines in quantities sufficient to produce symptoms (5,12,15,63,66,97,98). The vast majority of secreting tumors produce norepinephrine. Tumors producing serotonin, kallikrein, and histamine precursors have also been described, often causing carcinoid-like syndromes (15,99). Symptoms associated with secreting paragangliomas are the same as those seen with pheochromocytomas, namely excessive sweating, hypertension, tachycardia, nervousness, and weight loss (12,97). Breakdown products of norepinephrine, including metanephrine (normal: <1.3 mg in 24 hr urine), and vanillylmandelic acid (normal: 1.8–7.0 mg in 24 hr urine), should be measured in serum and urine (12). Catecholamine levels four to five times the upper limit of normal are required to produce clinical symptoms (15,97) and patients with secreting tumors frequently have levels elevated tenfold (12,97). In symptomatic patients, an abdominal CT and serum epinephrine assays should also be performed to rule out a pheochromocytoma.
Natural History of Paragangliomas
(A)
(B)
Figure 2 CT scan of a patient with a left jugular paraganglioma (A) with metastasis to a left level III lymph node (B). The axial CT of the skull base demonstrates bone destruction at the jugular foramen.
The natural history of paragangliomas is slow and steady growth. Tumors followed radiographically show an increase in size of less than 5 mm per year (100). Jansen et al. reviewed 48 patients with untreated head and neck paragangliomas, over a mean follow-up of 4.2 years (101). Sixty percent of tumors grew. The median increase in this subgroup of growing tumors was 1 mm/yr with a range of 0.3 to 5 mm/yr, and the median doubling time was 4.2 yrs (range 0.6–21.5). Overall, the median increase in all tumors was 0.83 mm/yr, and tumor doubling was 10 years. There was no significant difference found in growth rate between sporadic and hereditary tumors. Approximately 67% of the paragangliomas gave rise to generally mild signs or symptoms, most commonly related to their size and location. No cranial nerve (CN) palsies developed in their series. Conversely, van der May et al. have
Chapter 38: Paragangliomas of the Head and Neck Table 4
Fisch Classification of Temporal Bone Paragangliomas
Class A (glomus tympanicum) Class B (glomus hypotympanicum) Class C
C1 C2 C3 C4 Class D Figure 3 Shamblin classification for carotid body tumors. From left to right: Type 1 tumors are localized tumors, relatively free of involvement of vessel wall and easily resected; Type 2 tumors are partially adherent to surrounding vessels but do not encase the vessel; and Type 3 tumors intimately surround and encase the carotid, as well as regional nerves.
shown that if untreated, approximately 75% of patients with CBTs will eventually develop CN deficits (21), and Bradshaw et al. noted an 8% incidence of lower CN loss in patients with VPs managed with observation after a mean follow-up of 8.5 years (102).
Classification and Staging The main classification system for staging CBTs was described by Shamblin et al. in 1971 and was developed to grade the difficulty of resection (Fig. 3) (7,21). Four classification systems have been proposed for temporal bone paragangliomas to aid surgical planning and to standardize reporting: the Fisch classification (Table 3) (45), the Glasscock–Jackson classification system (Table 4) (40,103), De La Cruz classification (Table 5) (44), and the McCabe/Fletcher Classification (Table 6) (104).
Diagnosis and Investigations Diagnosis of paragangliomas rests on clinical examination and characteristic findings on diagnostic imaging. Symptoms suggestive of a functioning tumor, as well as a complete famTable 3 Glasscock–Jackson Classification of Glomus Tympanicum and Glomus Jugulare Glomus tympanicum I. Small mass limited to the promontory II. Tumor completely filling middle ear space III. Tumor filling middle ear and extending into the mastoid IV. Tumor filling middle ear, extending into mastoid or through the tympanic membrane to fill the external auditory canal, may extend anterior to the carotid Glomus jugulare I. Small tumor involving the jugular bulb, middle ear and mastoid II. Tumor extending under internal auditory canal; may have internal extension III. Tumor extending into the petrous apex, may have intracranial extension IV. Tumor extending beyond the petrous apex into clivus or infratemporal fossa; may have intracranial extension
543
De1 De2 Di1 Di2
Limited to mesotympanum Limited to the hypotympanum, mesotympanum, and mastoid without erosion of the jugular bulb Involvement and destruction of infralabyrinthine and apical compartments. Subclassification by degree of carotid canal erosion No invasion of the carotid; destruction of jugular bulb/foramen Invasion of vertical carotid canal between foramen and bend Invasion along horizontal carotid canal Invasion of foramen lacerum and along carotid into cavernous sinus Intracranial extension (De, extradural; Di, intradural) Up to 2 cm dural displacement More than 2 cm dural displacement Up to 2 cm intradural extension More than 2 cm intradural extension
ily history should be elicited on history. Diagnostic imaging should be performed with a suspicion of a paraganglioma. While fine-needle aspiration can safely be performed, with minimal risk of bleeding (12), it is difficult to interpret and does not usually aid diagnosis (6,105). An incisional biopsy is contraindicated. Clinical and radiographic features of paragangliomas are usually distinctive enough for definitive management. In patients presenting with paragangliomas of temporal bone an audiogram should be obtained.
Clinical Presentation Carotid Body Tumors A CBT commonly presents as an asymptomatic, slowly enlarging pulsatile neck mass located below the angle of mandible, along the anterior border of the sternocleidomastoid muscle (Fig. 4) (106). Typically, they are mobile in the horizontal but not vertical plane, a finding known as Fontaine sign (106). Rarely, a bruit or thrill may be present on auscultation or palpation, suggesting significant arterial compression (51). Bulging of the oropharyngeal wall may be seen with parapharyngeal space extension (Fig. 5). Shamblin type II and III tumors make up the majority of tumors in most series (21,106–108). Ten percent of patients will present with a CN deficit; the vagus nerve is most commonly affected (21). The recurrent laryngeal nerve is involved in 8% of cases, the hypoglossal in 6%, and the sympathetic chain in 2% (21,100,109). Cranial nerve deficits at presentation are more frequently seen with large (>5 cm) compressive tumors (51,110). Table 5 Antonio De La Cruz Classification for Jugulotympanic Paragangliomas Anatomic classification
Surgical approach
Tympanic Tympanomastoid Jugular bulb
Transcanal Mastoid-extended facial recess Mastoid-neck (possible limited facial nerve rerouting) Infratemporal fossa Infratemporal fossa/intracranial
Carotid artery Transdural
544
Goldstein et al.
Table 6 McCabe/Fletcher Classification of Temporal Bone Paragangliomas Group 1 1. Absence of bone destruction 2. Intact eight nerves 3. Intact jugular foramen nerves 4. Absence of facial weakness Group 2 1. Bone destruction confined to mastoid 2. Facial nerve normal or paretic 3. Jugular foramen nerves intact 4. Superior bulb of jugular vein uninvolved ny retrograde jugulography Group 3 1. Destruction on roentenogram involving petrous bone, jugular fossa, and occipital bone 2. Positive retrograde jugulography 3. Jugular foramen syndrome 4. Presence of metastases 5. Carotid arteriogram evidence of destruction of pertrous/ occipital bone
Jugulotympanic Paragangliomas Hearing loss and pulsatile tinnitus are the most common presenting symptoms, manifesting in 55% to 58% and 56% to 82% of patients respectively (15,111–115). Hearing loss is typically conductive, but, with tumor invasion into the labyrinth, patients may develop a sensorineural hearing loss and vertigo (15). On examination a reddish-blue, pulsating mass may be seen behind the inferior portion of an intact tympanic membrane (Fig. 6); pneumatic otoscopy causes the mass to blanch (Brown sign). Tympanic paragangliomas tend to remain well-defined, intra-tympanic soft-tissue masses (Fig. 7). In contrast the presence of intracranial extension in jugular paragangliomas at presentation has been reported to be as high as 72% (45,46,116). Whereas tympanic paragangliomas commonly manifest themselves early with pulsatile tinnitus and conductive hearing loss, jugular paragangliomas typically present late, frequently with at least one CN neuropathy (111). Palsy of the lower CNs usually precludes the diagnosis of a tympanic paraganglioma (117); neuropathies in tympanic para-
Figure 4 Clinical appearance of a carotid body tumor presenting as a lateral neck mass anterior to the sternomastoid muscle at the level of the carotid bifurcation.
Figure 5 CT scan of a large left carotid body tumor extending into the parapharyngeal space with bulging of the oropharyngeal wall.
gangliomas occur uncommonly and are usually limited to the seventh and eighth CN (40). The reported rates of CN neuropathies in jugular paragangliomas ranges from 8% to 60% (38,45,112,115,118–121). Palsies of the ninth, tenth, and eleventh CNs, producing jugular foramen (Vernet’s) syndrome may be occasionally encountered (38). Further skullbase extension can result in hypoglossal nerve involvement
Figure 6 Otoscopic examination of a patient with a jugular paraganglioma demonstrating a reddish-blue mass behind the inferior portion of an intact tympanic membrane.
Chapter 38: Paragangliomas of the Head and Neck
545
presentation, with the vagus nerve most commonly affected (47,66,130,131). Miller et al. reported that at presentation at least 50% of patients will have 1 or more CN palsies (47). Netterville et al. noted that 36% of patients presented with CN deficits (66). In 28% of patients the vagus nerve was totally or partially paralyzed. Other CN deficits upon presentation included CN XII (17%), CN XI (11%), CN IX (11%), CN VII (6%) and sympathetic chain (4%) (66).
Laryngeal and Sinonasal Paragangliomas Patients with laryngeal paragangliomas commonly present with varying degrees of dysphonia, stridor, dysphagia, and dyspnea (16,51). Supraglottic paragangliomas cause hoarseness and impair deglutition, whereas subglottic tumors produce airway obstruction (16,132). Fiberoptic examination classically reveals a submucosal mass with normal vocal cord function (16). Sinonasal paragangliomas present in a manner similar to that of any expansile mass lesion of the nasal cavity with nasal obstruction and epistaxis (16,51). These tumors are usually painless and may be misdiagnosed as nasal polyps (51). With advanced disease local symptoms will result from involvement of the facial skin and/or orbit (16,51).
Diagnostic Imaging Figure 7 Otoscopic examination of a patient with a tympanic paraganglioma immediately inferior to the umbo of the malleus. Unlike a jugular paraganglioma the mass is situated over the promontory and well-circumscribed with all margins visible.
(Collet–Sicard syndrome). Facial nerve paralysis is associated with advanced disease (114) causing medial and posterior extension of the tumor through the facial recess and retrofacial air cell tract to involve the facial nerve in its horizontal and vertical segment (122). Horner syndrome may occur with involvement of the sympathetic plexus, particularly when the carotid artery is involved in its petrous or cervical portion (44). The range in percentages of CN deficits at presentation is presented in Table 7 (114,116,118,122–128). In a literature review of 509 cases of jugular paragangliomas, Lustig et al. reported the rate of preoperative paralysis for the following CNs to be: 19% for CN IX, 24% for CN X, 16% for CN XI, and 21% for CN XII (129).
Vagal Paragangliomas VPs most commonly present as a painless upper neck mass, which is typically located more cephalad than CBTs (130). Tumors originating from the nodose ganglion manifest clinically as a mass behind the angle of the mandible, with associated hoarseness and vocal cord paralysis. Those that arise more superiorly impinge on the jugular foramen, producing Vernet or Collet–Siccard syndrome (15). With temporal bone involvement, hearing loss and pulsatile tinnitus may be present (47,66). Overall, CN paralysis is common at Table 7 Rate of Cranial Nerve Deficits at Presentation of Jugulotympanic Paragangliomas Cranial nerve IX X XI XII
Preoperative paralysis rate 4–43% 2–57.1% 4–43% 7–57%
Source: From Refs. 114,116,118,122–128.
Diagnostic imaging is an integral aspect of the work-up of patients presenting with symptoms and signs suggestive of a paraganglioma. Imaging aids diagnosis and treatment by delineating tumor location, displacement of major vessels, involvement or invasion of surrounding structures, and detection of multicentric tumors and metastases.
Ultrasound Ultrasound (US) can be used in the initial assessment of a neck mass suspected to be a paraganglioma, and is often sufficient to make the preliminary diagnosis of a CBT. In B-mode US paragangliomas present as a solid, well-defined, hypoechoic, heterogeneous mass (133). Color Doppler US can define the vascularity of the mass and its relationship to the ICA and ECA133. Because most VPs extend above the hyoid bone, the value of US for these lesions becomes limited (134). The direction of blood flow can, however, help differentiate CBTs from VPs; CBTs exhibit upward intratumoral blood flow (133), while downward intratumoral flow of a similarly located neck mass suggests a VP (133).
Computed Tomography High-resolution CT scanning is used to evaluate most patients with suspected head and neck paragangliomas. CT shows the anatomic relationship of the tumors with surrounding structures, and their relationship with the major vessels of the neck. Accuracy diminishes with lesions less than 8 mm in size (12,76). In general, paragangliomas appear as homogenous masses with intense enhancement following contrast administration. CT can differentiate between CBTs and VPs based on the relationship to the ICA and ECA (Figs. 8 and 9) (15,134). CT is particularly valuable in assessing bone invasion at the skull base (Fig. 2), which aids in differentiating jugular and tympanic paragangliomas, as well as in staging these tumors. If CT demonstrates an intact bone plate separating a middle ear mass from the dome of the jugular bulb a tympanic paraganglioma is the most likely diagnosis (15). If the caroticojugular spine, which separates the petrous carotid artery from the jugular bulb, is destroyed, a jugular paraganglioma extending into middle ear is more likely (15). However, the corollary is not always true; its preservation
546
Goldstein et al.
Magnetic Resonance Imaging
Figure 8 CT demonstrating the typical relationship of a left carotid body tumor to the carotid artery with splaying of the bifurcation, and displacement of the ICA posterolaterally and the ECA anterolaterally or anteromedially.
does not necessarily preclude the existence of a jugular paraganglioma (43,135). Finally, CT may also help rule out other causes of pulsatile tinnitus, such as an aberrant carotid artery, a high jugular bulb, or a persistent stapedial artery.
Figure 9 CT of a patient with bilateral vagal paragangliomas demonstrating displacement of the ECA and ICA anteromedially (circled), without splaying at the carotid bifurcation.
MRI is complementary to CT in patients being investigated for paragangliomas. MRI offers several benefits including (a) detection of smaller paragangliomas (12,37,136–138) (b) increased sensitivity for invasion or erosion of adjacent vessels, dura, or brain (12), and (c) the ability of magnetic resonance angiography (MRA) and venography (MRV) to provide noninvasive delineation of adjacent arterial and venous structures. On T1- and proton-weighted images paragangliomas display a low-to-intermediate signal; they are relatively hyperintense on T2-weighted images (12,15). On postcontrast T1-weighted spin-echo images, the tumors enhance strongly and homogenously (15,134). Paragangliomas exhibit a characteristic “salt and pepper” pattern of hyper- and hypointensity on T1- and T2-weighted images, most often evident in tumors greater than 1.5 cm in diameter (15,136,138). This appearance, which represents the existence of sinusoidal flow-voids within the tumor, is accentuated with gadolinium (Fig. 10). While characteristic, these findings are not specific to paragangliomas and may be seen with other hypervascular lesions, such as metastatic renal cell carcinoma (134,139). Short T1-inversion recovery (STIR) sequences are now also commonly used to reduce signal from fat, aiding in lesion detection (15). MRA has the ability to facilitate detection of multicentric paragangliomas and to determine feeding vessels preoperatively (Fig. 11) (138,140). Three-dimensional time of flight (TOF) MRA allows detection of vessel displacement, gross tumor involvement, and compromised blood flow (12). The combined use of spin-echo images and 3D TOF MRA has a higher sensitivity and similar specificity (90%/92%) to conventional 3D phase contrast angiography (72%/97%) (37,124,138,141,142). MRV can also demonstrate compression or invasion of the jugular bulb (15). Even with the techniques
Figure 10 MRI of a left jugular paraganglioma, axial T1-weighted study with gadolinium demonstrating the classic “salt and pepper” pattern of hyperand hypointensity.
Chapter 38: Paragangliomas of the Head and Neck
547
Figure 12 Conventional angiography of a jugular paraganglioma demonstrating an intense tumor blush confirming the tumor’s vascularity. Feeding vessels can also be assessed. Figure 11 MRA of a carotid body tumor demonstrating the “lyre sign,” which results from splaying of the internal and external carotid arteries.
discussed, MRA is not sensitive enough to demonstrate the detailed vascular supply to these tumors, which are best evaluated with digital subtraction superselective angiography when required (140).
Diagnostic Angiography Digital subtraction angiography is able to delineate, in exquisite detail, the vascular anatomy of these tumors, their feeding blood supply, and any potential vascular invasion or compromise. It is a sensitive test for small and multicentric paragangliomas (19,74,118,143). The hallmark appearance of a carotid body tumor on angiography is the “lyre sign,” which results from splaying of the carotids, and an intense tumor blush confirming the tumor’s vascularity. Temporal bone paragangliomas also reveal a characteristic angiographic appearance, consisting of enlarged feeding arteries, an early, intense, and slightly inhomogeneous tumor blush, and early-appearing draining veins (Fig. 12) (15). Despite its benefits, angiography is an invasive procedure with potential risks and is rarely required for diagnosis, assessment, and treatment planning. Its use is therefore currently limited to cases in which the diagnosis remains in question after CT and MRI (15,144), cases in which preoperative embolization is desired, or for preoperative planning in cases when there is suspicion of carotid artery involvement. In the latter circumstance angiography allows for assessment of the adequacy of the patency of the intracranial circulation (134), and, when used in conjunction with balloon occlusion to assess the adequacy of contralateral cerebral blood flow, can guide decisions regarding carotid artery sacrifice or bypass.
Octreotide Scintigraphy Paragangliomas, as with other neuroendocrine tumors, express somatostatinergic receptors and are well suited for imaging with octreotide scintigraphy (5,145). Octreotide is a somatostatin analog that, when coupled to a radioisotope
(111 indium-labeled-DPTA), produces a scintigraphic image (134,146). Areas of increased uptake can then be examined further with MRI. Octreotide scintigraphy appears to be a safe and relatively noninvasive method for early diagnosis of familial paragangliomas and is useful for the detection of multicentric tumors (134,147). The reported sensitivity and specificity in diagnosing paragangliomas (usually >1 cm) ranges from 94% to 97% and 75% to 82%, respectively (146,148,149). It has also been shown to have a role in detecting recurrent paragangliomas since octreotide binding is not affected by postsurgical or radiation changes (145,149).
Management Treatment options in the management of head and neck paragangliomas include observation, surgery, radiation, singly or in combination. Traditionally, surgery has been the preferred modality of treatment (12), while radiation was reserved for unresectable tumors or those occurring in elderly or debilitated patients. Advances in radiation therapy have led to improved long-term response with acceptable complications (12), allowing it to be considered a primary treatment option (112). Decisions regarding treatment modality must take into consideration the biologic behavior, size, and site of the tumor, the age and general medical status of the patient, and potential treatment-related morbidity (51). An evidence-based approach to optimal treatment for patients with paragangliomas is limited by the rarity of the tumor and the nature of available studies. The vast majority of studies are retrospective with small subject numbers. They often include a heterogeneous population of tumors arising at different sites and treated in nonuniform ways over many decades, during which time there have been significant advances in interventional radiology, surgery, and radiotherapy. Direct comparison between surgery and radiation is made difficult by the fact that the two treatment modalities measure success by different standards: in general, surgical success is measured by total tumor resection without recurrence, whereas radiation therapy judges treatment response by the
548
Goldstein et al.
absence of radiologic evidence of tumor growth. In addition, there is often a selection bias in these studies, with more advanced tumors referred for radiation and smaller tumors approached surgically. Based on current literature control rates for surgery and radiation are similar. Treatment modality decisions are therefore often based primarily on the risk of treatment-related complications.
Observation Based on the natural history of these tumors observation may be sufficient in selected cases, such as asymptomatic elderly patients with jugular paragangliomas, and patients with multiple, small, synchronous, and asymptomatic tumors. Serial imaging should be performed on a yearly basis; if significant tumor growth or clinical symptoms are noted, consideration can then be given to other interventions.
Surgery Surgery should be considered when tumors can be completely excised without significant morbidity. This group is usually limited to patients with small to moderate size CBT and tympanic paragangliomas who present without cranial nerve dysfunction. Surgery is also indicated for patients with jugular or VPs who present with lower CN dysfunction, in whom tumor resection will not result in additional significant functional deficits. Relative contraindications to surgery include extensive skull base or intracranial involvement, advanced patient age, medical comorbidities, and bilateral or multiple paragangliomas, the resection of which may result in unacceptable postoperative morbidity and bilateral lower CN palsies (12). If surgery results in significant disability requiring long-term rehabilitation, radiation should be considered (112). Since paragangliomas are often situated near or involve the skull base and carotid artery, they are best managed by a team of head and neck surgeons, neurosurgeons, neuro-otologists, and vascular surgeons. Speech therapists and physiotherapists are also integral members of the multidisciplinary team, addressing the potential postoperative speech, swallowing, and shoulder difficulties that may arise. Appropriate preoperative management of secreting tumors is essential. In cases in which tumors being considered for surgery involve the carotid artery, consideration should also be given to assessment of cerebral blood flow and possible balloon occlusion.
Superselective Angiography and Preoperative Embolization There is considerable debate in the literature about the usefulness of preoperative embolization (39,67,106,150). Advantages include decreased blood loss, a lower transfusion requirement (38,118), and, consequently, a reduction in operative time (15). However, embolization of paragangliomas is associated with the risk of stroke or blindness due to potential anastomosis between the ICA and ECA. The decision to use preoperative embolization depends on tumor location and size. Embolization has a well-recognized role in the surgical management of jugular paragangliomas (118), whereas tympanic paragangliomas usually do not require it. The use of preoperative embolization in the treatment of CBTs and VPs is a matter for debate. When indicated, surgery should be performed within two days of embolization in order to avoid recruitment of collateral tumor blood supply in the postinflammatory phase (118). Intraoperative ligation
of the ECA must also be avoided, as doing so prevents future embolization for recurrences.
Surgery for Carotid Body Paragangliomas Surgery is the preferred modality for treatment of CBTs (6,7,21,108,110,124,151). Relative contraindications to surgery particular to CBTs include Shamblin type III tumors and patients with bilateral tumors who have sustained a cranial nerve or sympathetic trunk injury on one side (6). Dissection of the tumor off of the carotid artery can be performed in either a subadventitial or peri-adventitial plane (152). The former carries a risk of arterial wall injury, whereas dissection in the capsular or peri-adventitial plane may carry a lower risk of vessel injury (153). Complete resection can usually be accomplished in most tumors and nerves adherent to the tumor can often be mobilized without injury (21). The extent of surgery and the risk of complications, such as stroke, carotid artery, and CN injury, depend on the Shamblin classification (7). Type I tumors can often be removed with few complications and without the need for shunting or carotid reconstruction (6). For complete removal of type II tumors, carotid shunting may be required, and type III tumors may require carotid reconstruction (6). When indicated, vascular bypass and reconstruction also allows resection of tumors with a low incidence of stroke (51,154). Intraluminal shunting should be employed in circumstances in which the common or ICA must be sacrificed or repaired, in order to reduce the risk of stroke (153,155,156). Ligation of the carotid artery without shunting must be avoided as it carries up to a 66% chance of stroke (21). Even patients who pass balloon occlusion testing still run a 25% chance of delayed stroke after carotid artery ligation without shunting (21,157). In over 75% of cases CBTs can be resected without the need for arterial reconstruction (156) and is more common in class II and III tumors, malignant lesions, and tumors over 6 cm in size (38,108,154,158). The most common complications are stroke and CN injury (110). The incidence of stroke is less than 2% in the current literature (12,108,155,156). The reported incidence of carotid artery injury is between 2% to 12.5% (39,107,159–161). Rates of postoperative CN injury range between 0% to 71% (162), with the more recent studies reporting rates between 13% and 56% and the majority occurring with type II or III tumors (106–108,110,160). Nerves at risk of injury include the vagus nerve, the superior laryngeal nerve, the glossopharyngeal nerve, the hypoglossal nerve, and the marginal mandibular branch of the facial nerve, with the vagus nerve being the most commonly injured (110).
Management of Bilateral Carotid Body Tumors Patients with bilateral CBTs present a management dilemma due to the potential for bilateral vagus nerve injury and baroreflex failure syndrome. Denervation of both carotid baroreceptors results in tonic inhibition of the parasympathetic input to arterial blood pressure and unopposed sympathetic activity (163,164). Patients experience sudden hypertension associated with episodes of flushing, headache, diaphoresis, and emotional lability, followed by sharp fluctuations in arterial pressure and heart rate typically within 24 to 72 hours after surgery (165). Marked hypotension and bradycardia may also occur when the patients are drowsy or sedated (165,166). There appears to be a wide spectrum of symptom severity, and timing of onset and resolution between patients (163,167). While there is an apparent heterogeneous response to arterial baroreflex dysfunction, the normocapnic hypoxic drive is invariably lost (164). The
Chapter 38: Paragangliomas of the Head and Neck
long-term clinical course is also variable: some patients experience a lifetime of hypertension while others improve without medical control (167). Clonidine is an excellent agent since it acts as a central alpha-2 agonist, thereby decreasing circulating norepinephrine levels (163,165). Episodes of hypotension have been shown to respond to the administration of low dose corticosteroids (68,165,168). There is no standard management protocol for bilateral carotid tumors. The decision of which modality to use on each CBT depends on the tumor stage, morbidity associated with its management and the preoperative status of the function of the vagus nerve. Whichever treatment modality is chosen every effort must be made to preserve the function of at least one vagal nerve and its laryngeal branches. One tumor can be resected with the contralateral CBT managed with observation, surgery, or radiation. If surgery is considered for both sides, the smaller tumor should be removed first with assessment of the function of the vagal nerve postoperatively before undertaking resection of the contralateral tumor. In cases where there is either pre- or postoperative vagal nerve paralysis, the contralateral tumor should be managed conservatively with either observation or radiation. Another treatment approach would be radiation to the tumor on both sides. However, baroreflex failure can also be a long-term sequelae of radiation therapy (169,170).
Surgery for Vagal Paragangliomas VPs confined to the neck have been frequently managed with surgery (51). Relative contraindications include elderly patients and bilateral tumors, since surgery inevitably results in sacrifice of the vagus nerve (66,160,171). The majority of tumors can be removed with a lateral cervical approach, even in tumors greater than 5 cm (47). With large tumors extending up to the skullbase, a mandibulotomy may be required for either adequate exposure or vascular control at the skull base (51). When surgery is performed for tumors that traverse the skull base, a combined cervical and mastoid approach may be indicated in order to achieve safe and wide exposure. The major morbidity following surgery is related to postoperative CN dysfunction, which is significantly more common with VPs than for CBTs (51). While an attempt can be made to preserve the vagus nerve, excision of these tumors almost always requires its sacrifice (47,66,102). Nerves which are spared frequently do not recover function (47). As a result, postsurgical CN deficits should be the main consideration when formulating a treatment plan. Vagal nerve paralysis is frequently complicated by an associated hypoglossal and glossopharyngeal nerve paralysis adding additional morbidity (47,66,102). Surgical resection may be also associated with vascular injury, especially in larger tumors and those with skull base involvement (51), although less frequently than CBTs (12). Biller et al. noted that the risk of carotid injury increases with increasing tumor size, as tumors larger than 5 cm tend to cause significant displacement of the carotid artery at the skull base (172). As a result of this high potential for morbidity from CN neuropathies and vascular injury, patients should be selected carefully for surgical management of VPs. Other management options such as radiation or observation should be considered, particularly in asymptomatic patients.
Surgery for Glomus Tympanicum Because the rate of intra- and postoperative morbidity is low in these patients, surgical resection is the treatment of choice for tympanic paragangliomas confined to the middle ear or mastoid (Fisch Class A and B) (45,67). Tumors confined to the
549
middle ear, in which the entire margin is visible (Glasscock– Jackson type 1, Fisch Class A), can be removed through a transcanal approach. Laser-assisted excision with either a KTP or an Nd:Yag laser has also been described for these tumors (173–175). Large tumors or those whose margins cannot be visually confirmed usually require an extended facial recess approach (117). This approach allows assessment of the relationship of the tumor to the ossicular chain, tympanic membrane, labyrinth, facial nerve, and jugular bulb (117). Surgical management results in total tumor removal and long-term tumor control in the majority of patients (117). Recurrence rates after surgical resection ranges from 2.5% to 5% (117,176); postoperative residual tumor is the main risk factor for recurrent disease. Complications are uncommon with Fisch Class A and B tumors (176,177). With more extensive tumors, complications include CSF leak, stroke, bleeding, hearing loss (conductive and/or sensorineural), and facial paralysis.
Surgery for Jugular Paragangliomas Management of jugular paragangliomas remains controversial. Surgery is curative, offering the possibility of immediate and complete tumor elimination; there is, however, considerable risk of morbidity. Advances in skull-base and microsurgical techniques, neuromonitoring, embolization, and preoperative evaluation of cerebral blood flow has given surgeons the ability to resect the majority of tumors with fewer complications (178). It is, therefore, the extent of morbidity, mainly related to CN dysfunction, that determines the role of surgery in the management of tumors in this location. The surgical approach to jugular paragangliomas must be tailored based on the routes of extension of the tumor, which are highly variable. The approach must take into account the tumor size, extent of distal ICA control necessary, and must provide access to all tumor margins. A combined transmastoid and transcervical approach can be used for small jugular paragangliomas that do not involve the carotid artery or posterior cranial fossa (44). Cranial nerves can be preserved if they can be adequately separated from the tumor (44). For larger tumors impinging on the facial nerve but still not involving the carotid artery or posterior cranial fossa, a similar approach can be used with limited facial nerve rerouting (44). More extensive tumors will require additional exposure with an infratemporal approach, affording access to the posterior and middle cranial fossa and petrous carotid artery. In these cases, complete exposure of the vertical portion of petrous carotid requires rerouting of the facial nerve from the geniculate ganglion laterally. Surgical exposure of the ICA must guarantee its proximal and distal control as well as visualization of its entire circumference to enable its mobilization. For tumors abutting or involving the ICA, accurate preoperative assessment of the extent of involvement and of cross-perfusion from contralateral vessels is imperative for safe resection of this tumor (116). When there is potential for carotid sacrifice, consideration should be given to balloon occlusion or revascularization (179). Involvement of the brain or cavernous sinus often makes complete resection impossible. Embolization continues to have a role in the surgical management of large tumors involving the jugular foramen and skull base (180). Gross tumor removal is possible in 70% to 96% of cases (45,116,123,178,181). The extent of resectability depends upon the stage and complexity of the tumor. Complex tumors (i.e., giant tumors, multiple, malignant or secreting tumors, previous treatments) have a significantly lower rate of complete resection than their noncomplex counterparts (178). Long-term
550
Goldstein et al. Table 8 Rate of Postoperative New Cranial Nerve Deficits Following Resection of Jugulotympanic Paragangliomas Cranial nerve VII IX X XI XII
Range 4.4–11% 26–42% 13–28% 25–26% 12.5–21%
Source: From Ref. 178.
local control is achieved in 80% to 95% of cases in most surgical series, with surgical control in noncomplex tumors approaching 95% (45,67,116,123,126,178,179,182–184). The overall recurrence rate is 2% to 3% when total tumor removal is possible, with a mean time to recurrence of 82.8 months (45). Because recurrences after surgery have been documented up to 23 years later, lifelong follow-up is required (179). Bleeding, stroke, CSF leak, and CN dysfunction are the most significant complications. Total resection without CN injury is only achievable in 31% of patients (51). Table 8 shows the rate of development of new lower CN injuries in more recent surgical series of jugular paragangliomas resections as reported by Gottrified et al. (178). The lack of preoperative CN dysfunction does not correlate well with intraoperative neural involvement and therefore does not predict postoperative deficits (116). Serviceable hearing is salvageable in over 90% of patients (126,178). Facial nerve injury most commonly results from manipulation of the nerve for access. Rates of recovery to House-Brackmann grade 1 or 2 levels of function vary from 60% to 87% depending on extent of rerouting (123,183–186). In most series of noncomplex tumors, the rate of CSF leak ranges from 3% to 11%, occurring most frequently in Fisch class D tumors (45,116,123,178,179,183). The majority resolve without surgical intervention (123,126,178). On a review of 384 patients from seven surgical series on jugular paragangliomas, Gottfried et al. noted the following complication rates: CSF leak occurred in 8.3% of patients, aspiration in 5.5%, and stroke in 1.6% (178). The overall mortality rate was 1.3%. Death usually resulted from injury to the ICA, intracerebral hemorrhage, or pulmonary embolism. In an attempt to reduce complications associated with intradural (Fisch class D2) tumors, some authors recommend staging the procedure, removing the extradural tumor first, followed by removal of the intradural at a later date (116,123). Rehabilitation is imperative in patients developing CN dysfunction following surgical resection of jugular paragangliomas. Full rehabilitation following surgical resection of class C and D tumors may take up to 1 to 2 years (181). With time and aggressive postoperative speech and swallowing therapy, patients may be able to compensate well (126,128,181). This, however, depends upon age, the CNs involved, and the preoperative status of the CNs. Younger patients and those with preoperative dysfunction have better outcomes with rehabilitation (126,187). There is wide variation in the role surgery plays in the management of jugular paragangliomas. Some authors suggest that gross total tumor removal should be performed in all young patients since they have a greater capacity to compensate for loss of CN function. In patients in whom tumor cannot be resected completely, subtotal resection followed by observation with serial imaging has been suggested as an option, since growth after surgical devascularization is rare; others follow with stereotactic radiation (116,185). Overall, surgery
can be considered in young patients with pre-existing CN dysfunction, in whom surgery is unlikely to cause additional morbidity, and it should generally not be offered in patients older than 60 years when preoperative CN deficits are not present. For all other tumors considerable debate still exists.
Radiation Advances in radiation therapy have led to improved longterm response with minimal morbidity (12). Currently, radiation therapy may be considered a reasonable primary treatment option for paragangliomas, particularly those in elderly patients, patients with multiple or severe medical conditions, or patients with extensive skull base or intracranial involvement (112). Radiotherapy is also indicated in patients with jugular or vagal paragangliomas without evidence of lower CN dysfunction or patients with multiple or bilateral tumors, which if surgically resected would result in significant morbidity and disability (12,68,120,171). Both conventional fractionated radiotherapy and stereotactic radiosurgery have been used to treat paragangliomas. Unlike with surgery, successful treatment with radiotherapy is defined as stability or regression of tumor size and neurological symptoms, not gross tumor removal (112). The exact mechanism by which radiation prevents tumor growth is not completely understood, although it is theorized that radiation induces an obliterative endarteritis with resultant fibrosis (112,178,188). It is difficult, however, to determine if radiation has a primary effect on the chief cells themselves: because morphologically intact but persistent cells may lose their ability to reproduce (189).
Conventional Fractionated Radiotherapy The doses of conventional radiotherapy commonly used in the literature ranges between 35 to 60 Gy. With these doses the majority of lesions remain stable in size or show modest regression on radiographic evaluation after completion of radiation (115,190,191). In general a total dose of 45 Gy in 25 fractions, using once-daily fractions, in continuous course delivered over 5 weeks is recommended to control benign paragangliomas (112,192,193). A schedule of 35 Gy delivered over 3 weeks has also been shown to be effective (113). Patients with unilateral disease can almost always be treated with an ipsilateral field arrangement, allowing for the minimization of long-term xerostomia. Patients with bilateral tumors are usually treated with more complex techniques. For malignant paragangliomas treated with surgery, postoperative radiation is often indicated at a more intensive dose fractionation schedule, typically 60 to 66 Gy in 6 to 6.5 weeks. There is an extensive body of literature with over 30 years of follow-up assessing the effectiveness of radiation therapy in controlling head and neck paragangliomas (112). Similar to the surgical literature, most of the radiation studies are single-institution, retrospective case series. There is significant heterogeneity both within and between studies in terms of years of analysis, dose levels, radiation source (electrons vs. photons), mode of delivery (external beam, stereotactic radiosurgery), and beam energies (superficial, cobalt, and 4–6 MV) (112). Despite these limitations, long-term control rates have been reported with radiation therapy (19). A review, by Hu et al., of 1000 tumors from 35 studies noted a 10-year local control rate for radiation therapy ranging from 65% to 100% with a mean of 90% (112). Schild et al. reported local control of 91% in a review of series published between 1965 to 1992 (194). Springate et al. also reviewed the literature on the management of head and neck paragangliomas published from 1965 to 1988; they found a control rate of 93%
Chapter 38: Paragangliomas of the Head and Neck
in patients treated with radiotherapy alone (195). It is worth noting, however, that studies reviewed often include patients managed with combined modality therapy. Radiation therapy has also been shown to result in resolution of symptoms and CN neuropathies in some patients. Tinnitus may improve or resolve in up to 88% of patients (112), and rates of improvement of CN neuropathies vary from 0% to 83%, with an average of 35% (112). Complete resolution of CN dysfunction occurs in about 10% of cases (range 8–20%). Notably about 1% of CN neuropathies worsen after radiotherapy (112,196,197). There is a growing body of literature to suggest that CBTs and VPs can be treated with radiation therapy when surgery will be associated with significant morbidity (6,112,162). Based on observational data, with limited numbers of patients, the overall likelihood of local control after radiation for cervical paragangliomas is essentially the same as surgery and complications are infrequent (104,192,198). While local control in a number of small series ranges from 96% to 100% and complications are rare (104,196,198–200), surgery is still considered the primary treatment for CBTs since the majority can be resected without significant morbidity. The most significant role for radiation is in the management of temporal bone paragangliomas, or VPs in which surgical removal can result in injury to the lower CNs and attendant significant morbidity. Hinerman et al. treated 55 temporal bone paragangliomas (46 JP, 9 GT) with radiation alone (104). Local control was obtained in 40 of 43 (93%) of previously untreated lesions. All Macabe-Fletcher Group I and II patients were controlled as opposed to a control rate of 87% for Group II tumors. The overall cause-specific survival was 98%, and no patients developed regional or distant metastasis. Two patients died, one of whom was treated with radiation for recurrence after surgery, and one of whom discontinued treatment. In the absence of tumor progression, no patients developed CN deficits. Local control rates with radiation alone for temporal bone paragangliomas range from 90% to 100% (43,104,113,195,201–204). Although some authors voice concern that tumor growth may resume 10 to 15 years later (43), literature supporting this notion is limited. Pemberton et al. assessed the management of glomus jugulare and glomus tympanicum tumors with external beam radiation therapy in 49 patients given a median dose of 45 Gy (201). At 5 and 10 years, 92% of patients were recurrence-free, and diseasespecific survival was 96%.
Stereotactic Radiosurgery First described in the mid-1990s (205,206), stereotactic radiosurgery or hyperfractionated stereotactic radiation has been used with increasing frequency in the treatment of paragangliomas of the head and neck. The initial small studies focused on jugular and tympanic tumors, and had limited follow-up times, but their results were promising enough to engender interest in this treatment modality. The earliest multicenter trial in jugular paragangliomas demonstrated that, in the majority of cases, stereotactic radiosurgery arrested tumor growth in 60% (28 of 47 patients) of patients and provided actual tumor regression in 40% (19/47). Symptoms improved in approximately 30% of patients, remained stable in 65% of patients, and deteriorated in approximately 5% of patients (207). Unfortunately, this trial was also limited in follow-up and utilized stereotactic radiosurgery both as a primary and secondary treatment modality, a difficulty that has plagued studies.
551
Further trials in jugular paragangliomas, with followup ranging from 2 to 7.2 years, have corroborated these findings. On average, tumors regress in 10% to 67% of patients and increase in size in only 5%. Subjective symptoms improve in up to 67% of patients, remain stable in approximately 30% to 40%, and increase in fewer than 10% (208–211). In single fraction schedules, dosages range from 10 to 30 Gy to the tumor margin, with a mean dose in most studies at approximately 15 Gy (207,209,212). As with the initial trials, however, most of these studies do not differentiate between radiosurgery as primary or secondary treatment. A review of published data comparing surgery with stereotactic radiosurgery as the primary treatment modality has found a similar recurrence rate (2.1% vs. 3.1%) between the two modalities, a similar rate of morbidity (8.3% vs. 8.5%), but an increased mortality rate (0% vs. 1.3%) with patients undergoing primary surgical treatment (178). However, in 92% of surgical cases, complete tumor removal was possible, whereas 100% of patients undergoing radiotherapy had residual tumor. Mean follow-up was approximately 4 years in this review, leading the authors to conclude that long-term control rates from stereotactic radiosurgery are still in question (178). In the single retrospective study with adequate followup (mean 13 years) (213), however, the 10-year tumor control rate has been found to be as high as 92%. This study followed 33 patients treated with either conventional (76%) or stereotactic (24%) radiotherapy. Nineteen patients (56%) had jugular paragangliomas, eight (24%) had tympanic tumors, and the remainder had carotid body or retroperitoneal tumors. In half the patients, radiotherapy was the primary modality. In one patient with a jugular paraganglioma, the tumor began to re-grow after 8.5 years; this patient was successfully treated with re-irradiation.
Complications Severe complications—defined as those requiring surgical intervention, hospitalization, hyperbaric oxygen, or death— are uncommon, and the majority of side effects are acute and mild, including mucositis, acute dermatitis, alopecia, otitis externa, serous otitis media, dysgeusia, and xerostomia (209,214). Severe complications reported include osteoradionecrosis of the temporal bone (1.7%), brain necrosis (0.84%), and radiation-induced malignant disease (0.28%) (19). Cranial neuropathies are also a potential risk of radiation, with up to 4% of patients undergoing radiotherapy for paragangliomas of the temporal bone developing facial nerve palsy (197). However, both osteoradionecrosis and cranial neuropathies are associated with doses of above 50 Gy (215). Cummings reported two major complications in 45 patients (4%): temporal bone necrosis after 70 Gy using electron beam and fatal brain necrosis after delivery of 70 Gy in 3 weeks in error (113). It should be noted that a number of authors have reported no complications in the management of CBT and VP with radiation (192,198,214,216). The risk of malignant transformation of the radiated paraganglioma, the development of radiation-induced malignancies, and difficulty with surgical salvage are all possible disadvantages of radiation therapy. The former two concerns, however, are not necessarily supported by evidence. Krych et al. reported on long-term outcomes following radiation for 33 paragangliomas treated over 28 years (25 with conventional radiotherapy and 8 with stereotactic radiosurgery) with a median follow-up of 161 months (range 4–429 mo) (213). No patient in their series developed a radiation-induced malignancy. The authors performed a MedLine search for second primary malignancies following radiation for
552
Goldstein et al.
paragangliomas and found only three reported cases that were possibly related to radiation (213). Lalwani et al. estimated the risk of radiation-induced malignancy to be 0.5–1 in 1000 (217), while others have placed the risk as high as 2 to 3 per 1000 (19,124,195,213). Complications from stereotactic radiosurgery so far have been rare. Isolated reports do exist of intractable vertigo (190,209) and new transient cranial nerve palsies (210), but no reports of permanent cranial nerve palsies or radiationinduced malignancies have been published, although current follow-up is still limited.
Comparison Between Radiation and Surgery Studies comparing the results of surgery and with those of radiation in the management of head and neck paragangliomas typically compare historical results published on each modality from retrospective case series. Selection bias also limits the interpretation of these results. Since we do not have randomized trials or well-designed observational studies directly comparing the two modalities in homogenous populations, one cannot determine if one treatment is superior to others. From the available literature, however, it appears that the rates of control after radiation or surgery are both around 90% (214). Therefore, treatment decisions must be based on the ability to control the tumor with minimal short- and longterm morbidity. Patient and physician preference both affect the choice of treatment. In most cases, the likelihood of cure after subtotal resection and radiation is similar to that after radiation alone; there therefore does not appear to be an advantage of planned combined therapy (193,197,214). Hu et al. performed a combined analysis of five studies which reported on results of surgery or radiation where tumors were staged using the McCabe/Fletcher system. Ninety-four patients were treated with radiation alone, 45 with surgery alone, and 39 with surgery and radiation. The average local control rates were 93%, 78%, and 85% for the three groups respectively. Sixty-five percent of tumors managed with radiation were stage III compared with 13% of tumors managed with surgery alone, and 51% of those treated with a combined approach (112). For recurrent chemodectomas, radiotherapy may be a better treatment modality than surgery. In a review of 29 patients with recurrent jugular and tympanic tumors (218), patients treated with stereotactic, conventional, or intensitymodulated radiotherapy fared better in terms of long-term disease-free survival (100% vs. 62% at 5 years) and posttreatment side effects (0% vs. 47%) than patients who were treated with a re-operation. All recurrent carotid body tumors, on the other hand, were successfully treated surgically.
CONCLUSION In general, carotid body tumors and low vagal paragangliomas can be treated surgically. Management of jugulotympanic tumor is, however, more controversial. Small tympanic tumors can be resected easily and with minimal morbidity, while large tumors require lateral skull base or infratemporal fossa approaches that are associated with morbidity; radiation therapy is often therefore considered. Radiation therapy is also advocated for patients with multiple paragangliomas, in patients with poor performance status or in elderly patients in whom disability from cranial nerve injury would be poorly tolerated. Observation may also be an option in the latter two groups of patients.
ACKNOWLEDGMENT We thank Cheryl Volling for drawing Shamblin classification.
REFERENCES 1. Lack EE, Cubilla AL, Woodruff JM, Farr HW. Paragangliomas of the head and neck region: A clinical study of 69 patients. Cancer. 1977;39:397–409. 2. Lee JH, Barich F, Karnell LH, et al. National Cancer Data Base report on malignant paragangliomas of the head and neck. Cancer. 2002;94:730–737. 3. Heth J. The basic science of glomus jugulare tumors. Neurosurg Focus 2004;17:E2. 4. Glenner GG, Grimley PM. Tumors of the extra-adrenal paraganglion system (including chemoreceptors). In: Atlas of Tumor Pathology. Washington, DC: Armed Forces Institute of Pathology, 1974:1–90. 5. Rao AB, Koeller KK, Adair CF. From the archives of the AFIP. Paragangliomas of the head and neck: radiologic-pathologic correlation. Armed Forces Institute of Pathology. Radiographics. 1999;19:1605–1632. 6. Knight TT, Jr., Gonzalez JA, Rary JM, et al. Current concepts for the surgical management of carotid body tumor. Am J Surg. 2006;191:104–110. 7. Shamblin WR, ReMine WH, Sheps SG, et al. Carotid body tumor (chemodectoma). Clinicopathologic analysis of ninety cases. Am J Surg. 1971;122:732–739. 8. Rosenwasser H. Carotid body tumor of the middle ear and mastoid. Arch Otolaryngol. 1945;41:64–67. 9. Guild SR. A hitherto unrecognized structure, the glomus jugularies in man. Anat Rec. 1941;79:28. 10. Gulya AJ. The glomus tumor and its biology. Laryngoscope. 1993;103:7–15. 11. Martin TP. What we call them: the nomenclature of head and neck paragangliomas. Clin Otolaryngol. 2006;31:185– 186. 12. Hu K, Persky MS. Multidisciplinary management of paragangliomas of the head and neck, Part 1. Oncology. (Williston Park) 2003;17:983–993. 13. Khan Q, Heath D, Smith P. Anatomical variations in human carotid bodies. J Clin Pathol. 1988;41:1196–1199. 14. Smith P, Jago R, Heath D. Anatomical variation and quantitative histology of the normal and enlarged carotid body. J Pathol. 1982;137:287–304. 15. Noujaim SE, Pattekar MA, Cacciarelli A, et al. Paraganglioma of the temporal bone: role of magnetic resonance imaging versus computed tomography. Top Magn Reson Imaging. 2000;11:108– 122. 16. Myssiorek D, Halaas Y, Silver C. Laryngeal and sinonasal paragangliomas. Otolaryngol Clin North Am. 2001;34:971–982, vii. 17. Wasserman PG, Savargaonkar P. Paragangliomas: Classification, pathology, and differential diagnosis. Otolaryngol Clin North Am. 2001;34:845–862, v–vi. 18. Myssiorek D. Head and neck paragangliomas: An overview. Otolaryngol Clin North Am. 2001;34:829–836, v. 19. Gujrathi CS, Donald PJ. Current trends in the diagnosis and management of head and neck paragangliomas. Curr Opin Otolaryngol Head Neck Surg. 2005;13:339–342. 20. McCaffrey TV, Myssiorek D, Marrinan M. Head and neck paragangliomas: physiology and biochemistry. Otolaryngol Clin North Am. 2001;34:837–844, v. 21. van der Mey AG, Jansen JC, van Baalen JM. Management of carotid body tumors. Otolaryngol Clin North Am. 2001;34:907– 924, vi. 22. Wenig B. Atlas of Head and Neck Pathology. Philadelphia: W.B. Saunders Company, 1993. 23. Kimura N, Chetty R, Capella C, et al. Extra-adrenal paraganglioma: Carotid body, jugulotympanic, vagal, laryngeal, aorticopulmonary. In: DeLellis RA LR, Heitz PU, Eng C, eds. World Health Organiztion: Pathology and Genetics: Tumors of Endocrine Organs. Lyon: IARC Press, 2004.
Chapter 38: Paragangliomas of the Head and Neck 24. Kliewer KE, Wen DR, Cancilla PA, et al. Paragangliomas: assessment of prognosis by histologic, immunohistochemical, and ultrastructural techniques. Hum Pathol. 1989;20:29–39. 25. Arts HA, Fagan PA. Vagal body tumors. Otolaryngol Head Neck Surg. 1991;105:78–85. 26. Michaelson PG, Fowler CB, Brennan J. Tracheal paraganglioma presenting with acute airway obstruction. Otolaryngol Head Neck Surg. 2005;132:661–662. 27. Del Gaudio JM, Muller S. Diagnosis and treatment of supraglottic laryngeal paraganglioma: report of a case. Head Neck. 2004;26:94–98. 28. Peterson KL, Fu YS, Calcaterra T. Subglottic paraganglioma. Head Neck. 1997;19:54–56. 29. Sanders KW, Abreo F, Rivera E, et al. A diagnostic and therapeutic approach to paragangliomas of the larynx. Arch Otolaryngol Head Neck Surg. 2001;127:565–569. 30. Ketabchi S, Massi D, Santoro R, et al. Paraganglioma of the nasal cavity: A case report. Eur Arch Otorhinolaryngol. 2003;260:336– 340. 31. Lecanu JB, Arkwright S, Halimi PH, et al. Multifocal malignant paraganglioma of the paranasal sinuses: A case report. Otolaryngol Head Neck Surg. 2002;126:445–447. 32. Mouadeb DA, Chandra RK, Kennedy DW, et al. Sinonasal paraganglioma: Endoscopic resection with a 4-year follow-up. Head Neck. 2003;25:1077–1081. 33. Sharma MCMC, Epari SS, Gaikwad SS, et al. Orbital paraganglioma: Report of a rare case. Can J Ophthalmol. 2005;40:640– 644. 34. Moyer JS, Bradford CR. Sympathetic paraganglioma as an unusual cause of Horner’s syndrome. Head Neck. 2001;23:338–342. 35. Kronz JD, Argani P, Udelsman R, et al. Paraganglioma of the thyroid: Two cases that clarify and expand the clinical spectrum. Head Neck. 2000;22:621–625. 36. Maxwell JG, Jones SW, Wilson E, et al. Carotid body tumor excisions: adverse outcomes of adding carotid endarterectomy. J Am Coll Surg. 2004;198:36–41. 37. van den Berg R, van Gils AP, Wasser MN. Imaging of head and neck paragangliomas with three-dimensional time-of-flight MR angiography. Am J Roentgenol. 1999;172:1667–1673. 38. Persky MS, Hu KS, Berenstein A. Paragangliomas of the head and neck In: Harrison LB, Sessions RB, Hong WK, eds. Head and Neck Cancer: A Multidisciplinary Approach. Philadelphia: Lippincott Williams & Wilkins, 2004:678–713. 39. Leonetti JP, Donzelli JJ, Littooy FN, et al. Perioperative strategies in the management of carotid body tumors. Otolaryngol Head Neck Surg. 1997;117:111–115. 40. Larson TC 3rd, Reese DF, Baker HL Jr., et al. Glomus tympanicum chemodectomas: Radiographic and clinical characteristics. Radiology. 1987;163:801–806. 41. Hesselink JR, Davis KR, Taveras JM. Selective arteriography of glomus tympanicum and jugulare tumors: Techniques, normal and pathologic arterial anatomy. Am J Neuroradiol. 1981;2:289– 297. 42. Mafee MF, Kumar A, Yannias DA, et al. Computed tomography of the middle ear in the evaluation of cholesteatomas and other soft-tissue masses: Comparison with pluridirectional tomography. Radiology. 1983;148:465–472. 43. Arriaga M. Paraganglioma (Glomus tympanicum). In: Jackler RK, Driscoll CLW, eds. Tumors of the Ear and Temporal Bone. Philadelphia: Lippincott Williams & Wilkins, 2000:112–127. 44. Friedman R, Brackmann DE. Jugulotympanic paragangliomas. In: Jackler R, Driscoll CLW, eds. Tumors of the Ear and Temporal Bone. Philadelphia: Lippincott Williams & Wilkins, 2000:343– 360. 45. Moe KS, Li D, Linder TE, et al. An update on the surgical treatment of temporal bone paraganglioma. Skull Base Surg. 1999;9:185–194. 46. Rigby P, Jackler RK. Clinicopathologic presentation and diagnostic imaging of jugular foramen tumors. Operat Tech Otolaryngol Head Neck Surg. 1996;7:99–105. 47. Miller RB, Boon MS, Atkins JP, et al. Vagal paraganglioma: The Jefferson experience. Otolaryngol Head Neck Surg. 2000;122:482–487.
553
48. Hinojar AG, Prieto JR, Munoz E, et al. Relapsing paraganglioma of the inferior laryngeal paraganglion: Case report and review of the literature. Head Neck. 2002;24:95–102. 49. Kuhn JA, Aronoff BL. Nasal and nasopharyngeal paraganglioma. J Surg Oncol 1989;40:38–45. 50. Lack E. Tumors of the adrenal gland and extra-adrenal paraganglioma. Atlas of Tumor Pathology, series 3, fasc 19. Washington, DC: Armed Forces Institute of Pathology, 1997. 51. Pellitteri PK, Rinaldo A, Myssiorek D, et al. Paragangliomas of the head and neck. Oral Oncol. 2004;40:563–575. 52. Astrom KK, Cohen JEJE, Willett-Brozick JEJE, et al. Altitude is a phenotypic modifier in hereditary paraganglioma type 1: Evidence for an oxygen-sensing defect. Human Genet. 2003;113:228–237. 53. Nissenblatt MJ. Cyanotic heart disease: “low altitude” risk for carotid body tumor? Johns Hopkins Med J. 1978;142:18–22. 54. Rodriguez-Cuevas HH, Lau II, Rodriguez HHP. High-altitude paragangliomas diagnostic and therapeutic considerations. Cancer. 1986;57:672–676. ´ 55. Rodr´ıguez-Cuevas SS, Lopez-Garza JJ, Labastida-Almendaro SS. Carotid body tumors in inhabitants of altitudes higher than 2000 meters above sea level. Head Neck. 1998;20:374–378. 56. Saldana MMJ, Salem LLE, Travezan RR. High altitude hypoxia and chemodectomas. Human Pathol. 1973;4:251–263. 57. Dundee P, Clancy B, Wagstaff S, et al. Paraganglioma: The role of genetic counselling and radiological screening. J Clin Neurosci. 2005;12:464–466. 58. Petropoulos AE, Luetje CM, Camarata PJ, et al. Genetic analysis in the diagnosis of familial paragangliomas. Laryngoscope 2000;110:1225–1229. 59. Drovdlic CM, Myers EN, Peters JA, et al. Proportion of heritable paraganglioma cases and associated clinical characteristics. Laryngoscope. 2001;111:1822–1827. 60. Grufferman S, Gillman MW, Pasternak LR, et al. Familial carotid body tumors: Case report and epidemiologic review. Cancer. 1980;46:2116–2122. 61. Parry DM, Li FP, Strong LC, et al. Carotid body tumors in humans: Genetics and epidemiology. J Natl Cancer Inst. 1982;68:573–578. 62. van der Mey AG, Maaswinkel-Mooy PD, Cornelisse CJ, et al. Genomic imprinting in hereditary glomus tumours: Evidence for new genetic theory. Lancet. 1989;2:1291–1294. 63. Badenhop RF, Jansen JC, Fagan PA, et al. The prevalence of SDHB, SDHC, and SDHD mutations in patients with head and neck paraganglioma and association of mutations with clinical features. J Med Genet. 2004;41:e99. 64. Spector GJ, Ciralsky R, Maisel RH, et al. Multiple glomus tumors in the head and neck. Laryngoscope 1975;85:1066–1075. 65. Hodge KM, Byers RM, Peters LJ. Paragangliomas of the head and neck. Arch Otolaryngol Head Neck Surg. 1988;114:872–877. 66. Netterville JL, Jackson CG, Miller FR, et al. Vagal paraganglioma: A review of 46 patients treated during a 20-year period. Arch Otolaryngol Head Neck Surg. 1998;124:1133–1140. 67. Boedeker CC, Ridder GJ, Schipper J. Paragangliomas of the head and neck: Diagnosis and treatment. Fam Cancer. 2005;4:55– 59. 68. Netterville JL, Reilly KM, Robertson D, et al. Carotid body tumors: A review of 30 patients with 46 tumors. Laryngoscope. 1995;105:115–126. 69. Carney JA, Sheps SG, Go VL, et al. The triad of gastric leiomyosarcoma, functioning extra-adrenal paraganglioma and pulmonary chondroma. N Engl J Med. 1977;296:1517–1518. 70. Hull MT, Roth LM, Glover JL, et al. Metastatic carotid body paraganglioma in von Hippel-Lindau disease. An electron microscopic study. Arch Pathol Lab Med. 1982;106:235–239. 71. Benn DE, Gimenez-Roqueplo AP, Reilly JR, et al. Clinical presentation and penetrance of pheochromocytoma/paraganglioma syndromes. J Clin Endocrinol Metab. 2006;91:827–836. 72. Neumann HP, Bausch B, McWhinney SR, et al. Germ-line mutations in nonsyndromic pheochromocytoma. N Engl J Med. 2002;346:1459–1466. 73. Baysal BE. Genetics of familial paragangliomas: Past, present, and future. Otolaryngol Clin North Am. 2001;34:863–879, vi.
554
Goldstein et al.
74. Maier W, Marangos N, Laszig R. Paraganglioma as a systemic syndrome: Pitfalls and strategies. J Laryngol Otol. 1999;113:978– 982. 75. Schimke RN, Collins DL, Rothberg PG. Functioning carotid paraganglioma in the von Hippel-Lindau syndrome. Am J Med Genet. 1998;80:533–534. 76. van Gils AP, van der Mey AG, Hoogma RP, et al. MRI screening of kindred at risk of developing paragangliomas: Support for genomic imprinting in hereditary glomus tumours. Br J Cancer. 1992;65:903–907. 77. Pawlu C, Bausch B, Neumann HP. Mutations of the SDHB and SDHD genes. Fam Cancer. 2005;4:49–54. 78. Baysal BE. Hereditary paraganglioma targets diverse paraganglia. J Med Genet. 2002;39:617–622. 79. Baysal BE, Ferrell RE, Willett-Brozick JE, et al. Mutations in SDHD, a mitochondrial complex II gene, in hereditary paraganglioma. Science. 2000;287:848–851. 80. Mariman EC, van Beersum SE, Cremers CW, et al. Fine mapping of a putatively imprinted gene for familial non-chromaffin paragangliomas to chromosome 11q13.1: Evidence for genetic heterogeneity. Hum Genet. 1995;95:56–62. 81. Mhatre AN, Li Y, Feng L, et al. SDHB, SDHC, and SDHD mutation screen in sporadic and familial head and neck paragangliomas. Clin Genet. 2004;66:461–466. 82. Baysal BE, Willett-Brozick JE, Lawrence EC, et al. Prevalence of SDHB, SDHC, and SDHD germline mutations in clinic patients with head and neck paragangliomas. J Med Genet. 2002;39:178– 183. 83. Dluhy RG. Pheochromocytoma—death of an axiom. N Engl J Med. 2002;346:1486–1488. 84. Schiavi F, Boedeker CC, Bausch B, et al. Predictors and prevalence of paraganglioma syndrome associated with mutations of the SDHC gene. JAMA. 2005;294:2057–2063. 85. Neumann HP, Pawlu C, Peczkowska M, et al. Distinct clinical features of paraganglioma syndromes associated with SDHB and SDHD gene mutations. JAMA. 2004;292:943–951. 86. Gimenez-Roqueplo AP, Favier J, Rustin P, et al. Mutations in the SDHB gene are associated with extra-adrenal and/or malignant phaeochromocytomas. Cancer Res. 2003;63:5615– 5621. 87. Young AL, Baysal BE, Deb A, et al. Familial malignant catecholamine-secreting paraganglioma with prolonged survival associated with mutation in the succinate dehydrogenase B gene. J Clin Endocrinol Metab. 2002;87:4101–4105. 88. Gimenez-Roqueplo AP, Favier J, Rustin P, et al. Functional consequences of a SDHB gene mutation in an apparently sporadic pheochromocytoma. J Clin Endocrinol Metab. 2002;87:4771– 4774. 89. Bikhazi PH, Roeder E, Attaie A, et al. Familial paragangliomas: The emerging impact of molecular genetics on evaluation and management. Am J Otol. 1999;20:639–643. 90. Renard L, Godfraind C, Boon LM, et al. A novel mutation in the SDHD gene in a family with inherited paragangliomas— implications of genetic diagnosis for follow up and treatment. Head Neck. 2003;25:146–151. 91. Rinaldo A, Myssiorek D, Devaney KO, et al. Which paragangliomas of the head and neck have a higher rate of malignancy? Oral Oncol. 2004;40:458–460. 92. Walsh RM, Leen EJ, Gleeson MJ, et al. Malignant vagal paraganglioma. J Laryngol Otol. 1997;111:83–88. 93. Martin CE, Rosenfeld L, McSwain B. Carotid body tumors: A 16-year follow-up of seven malignant cases. South Med J. 1973;66:1236–1243. 94. Zbaren P, Lehmann W. Carotid body paraganglioma with metastases. Laryngoscope. 1985;95:450–454. 95. Sykes JM, Ossoff RH. Paragangliomas of the head and neck. Otolaryngol Clin North Am. 1986;19:755–767. 96. Manolidis S, Shohet JA, Jackson CG, et al. Malignant glomus tumors. Laryngoscope. 1999;109:30–34. 97. Schwaber MK, Glasscock ME, Nissen AJ, et al. Diagnosis and management of catecholamine secreting glomus tumors. Laryngoscope 1984;94:1008–1015.
98. Erickson D, Kudva YC, Ebersold MJ, et al. Benign paragangliomas: Clinical presentation and treatment outcomes in 236 patients. J Clin Endocrinol Metab. 2001;86:5210–5216. 99. Jensen NF. Glomus tumors of the head and neck: Anesthetic considerations. Anesth Analg. 1994;78:112–119. 100. van der Mey AG, Frijns JH, Cornelisse CJ, et al. Does intervention improve the natural course of glomus tumors? A series of 108 patients seen in a 32-year period. Ann Otol Rhinol Laryngol. 1992;101:635–642. 101. Jansen JC, van den Berg R, Kuiper A, et al. Estimation of growth rate in patients with head and neck paragangliomas influences the treatment proposal. Cancer. 2000;88:2811–2816. 102. Bradshaw JW, Jansen JC. Management of vagal paraganglioma: Is operative resection really the best option? Surgery. 2005;137:225–228. 103. Jackson CG, Glasscock ME, 3rd, Harris PF. Glomus tumors. Diagnosis, classification, and management of large lesions. Arch Otolaryngol. 1982;108:401–410. 104. Hinerman RW, Mendenhall WM, Amdur RJ, et al. Definitive radiotherapy in the management of chemodectomas arising in the temporal bone, carotid body, and glomus vagale. Head Neck. 2001;23:363–371. 105. Monabati A, Hodjati H, Kumar PV. Cytologic findings in carotid body tumors. Acta Cytol. 2002;46:1101–1104. 106. Patetsios P, Gable DR, Garrett WV, et al. Management of carotid body paragangliomas and review of a 30-year experience. Ann Vasc Surg. 2002;16:331–338. 107. Luna-Ortiz K, Rascon-Ortiz M, Villavicencio-Valencia V, et al. Carotid body tumors: Review of a 20-year experience. Oral Oncol. 2005;41:56–61. 108. Plukker JT, Brongers EP, Vermey A, et al. Outcome of surgical treatment for carotid body paraganglioma. Br J Surg. 2001;88:1382–1386. 109. Nora JD, Hallett JW Jr., O’Brien PC, et al. Surgical resection of carotid body tumors: Long-term survival, recurrence, and metastasis. Mayo Clin Proc. 1988;63:348–352. 110. Najibi S, Terramani TT, Brinkman W, et al. Carotid body tumors. J Am Coll Surg. 2002;194:538–539. 111. Holden PK, Linthicum FH, Jr. Glomus jugulare tumor. Otol Neurotol. 2005;26:312–313. 112. Hu K, Persky MS. The multidisciplinary management of paragangliomas of the head and neck, Part 2. Oncology (Williston Park). 2003;17:1143–1153; discussion 1154, 1158, 1161. 113. Cummings BJ, Beale FA, Garrett PG, et al. The treatment of glomus tumors in the temporal bone by megavoltage radiation. Cancer. 1984;53:2635–2640. 114. Jackson CG. Glomus tympanicum and glomus jugulare tumors. Otolaryngol Clin North Am. 2001;34:941–970, vii. 115. Pryzant RM, Chou JL, Easley JD. Twenty year experience with radiation therapy for temporal bone chemodectomas. Int J Radiat Oncol Biol Phys. 1989;17:1303–1307. 116. Sanna MM, Jain YY, De Donato GG, et al. Management of jugular paragangliomas: The Gruppo Otologico experience. Otology Neurotol. 2004;25:797–804. 117. Forest JA, 3rd, Jackson CG, McGrew BM. Long-term control of surgically treated glomus tympanicum tumors. Otol Neurotol. 2001;22:232–236. 118. Persky MS, Setton A, Niimi Y, et al. Combined endovascular and surgical treatment of head and neck paragangliomas—a team approach. Head Neck. 2002;24:423–431. 119. Jackson CG, Harris PF, Glasscock ME, 3rd, et al. Diagnosis and management of paragangliomas of the skull base. Am J Surg. 1990;159:389–393. 120. Gardner G, Cocke EW, Jr., Robertson JH, et al. Skull base surgery for glomus jugulare tumors. Am J Otol. 1985;Suppl:126–134. 121. Ogura JH, Spector GJ, Gado M. Glomus jugulare and vagale. Ann Otol Rhinol Laryngol. 1978;87:622–629. 122. Spector GJ, Gado M, Ciralsky R, et al. Neurologic implications of glomus tumors in the head and neck. Laryngoscope. 1975;85:1387–1395. 123. Pareschi R, Righini S, Destito D, et al. Surgery of glomus jugulare tumors. Skull Base. 2003;13:149–157.
Chapter 38: Paragangliomas of the Head and Neck 124. Alford BR, Guilford FR. A comprehensive study of tumors of the glomus jugulare. Laryngoscope. 1962;72:765–805. 125. Makek M, Franklin DJ, Zhao JC, et al. Neural infiltration of glomus temporale tumors. Am J Otol. 1990;11:1–5. 126. Patel SJ, Sekhar LN, Cass SP, et al. Combined approaches for resection of extensive glomus jugulare tumors. A review of 12 cases. J Neurosurg. 1994;80:1026–1038. 127. Pensak ML, Jackler RK. Removal of jugular foramen tumors: The fallopian bridge technique. Otolaryngol Head Neck Surg. 1997;117:586–591. 128. Woods CI, Strasnick B, Jackson CG. Surgery for glomus tumors: The Otology Group experience. Laryngoscope. 1993;103:65– 70. 129. Lustig LLR, Jackler RRK. The variable relationship between the lower cranial nerves and jugular foramen tumors: Implications for neural preservation. Am J Otol. 1996;17:658–668. 130. Sniezek JC, Netterville JL, Sabri AN. Vagal paragangliomas. Otolaryngol Clin North Am. 2001;34:925–939, vi. 131. Lawson W. Glomus bodies and tumors. N Y State J Med. 1980;80:1567–1575. 132. Myssiorek D, Rinaldo A, Barnes L, et al. Laryngeal paraganglioma: An updated critical review. Acta Otolaryngol. 2004;124:995–999. 133. Stoeckli SJ, Schuknecht B, Alkadhi H, et al. Evaluation of paragangliomas presenting as a cervical mass on color-coded Doppler sonography. Laryngoscope. 2002;112:143–146. 134. Lustrin ES, Palestro C, Vaheesan K. Radiographic evaluation and assessment of paragangliomas. Otolaryngol Clin North Am. 2001;34:881–906, vi. 135. Phelps PD, Stansbie JM. Glomus jugulare or tympanicum? The role of CT and MR imaging with gadolinium DTPA. J Laryngol Otol. 1988;102:766–776. 136. Olsen WL, Dillon WP, Kelly WM, et al. MR imaging of paragangliomas. Am J Roentgenol. 1987;148:201–204. 137. Phelps PD, Cheesman AD. Imaging jugulotympanic glomus tumors. Arch Otolaryngol Head Neck Surg. 1990;116:940–945. 138. Vogl TJ, Juergens M, Balzer JO, et al. Glomus tumors of the skull base: Combined use of MR angiography and spin-echo imaging. Radiology. 1994;192:103–110. 139. Som PM, Braun IF, Shapiro MD, et al. Tumors of the parapharyngeal space and upper neck: MR imaging characteristics. Radiology. 1987;164:823–829. 140. van den Berg R, Wasser MN, van Gils AP, et al. Vascularization of head and neck paragangliomas: Comparison of three MR angiographic techniques with digital subtraction angiography. Am J Neuroradiol. 2000;21:162–170. 141. van den Berg R, Verbist BM, Mertens BJ, et al. Head and neck paragangliomas: Improved tumor detection using contrastenhanced 3D time-of-flight MR angiography as compared with fat-suppressed MR imaging techniques. Am J Neuroradiol. 2004;25:863–870. 142. van den Berg R, Schepers A, de Bruine FT, et al. The value of MR angiography techniques in the detection of head and neck paragangliomas. Eur J Radiol. 2004;52:240–245. 143. Wax MK, Briant TD. Carotid body tumors: A review. J Otolaryngol. 1992;21:277–285. 144. van den Berg R. Imaging and management of head and neck paragangliomas. Eur Radiol. 2005;15:1310–1318. 145. Duet M, Sauvaget E, Petelle B, et al. Clinical impact of somatostatin receptor scintigraphy in the management of paragangliomas of the head and neck. J Nucl Med. 2003;44:1767–1774. 146. Bustillo A, Telischi F, Weed D, et al. Octreotide scintigraphy in the head and neck. Laryngoscope. 2004;114:434–440. 147. Myssiorek D, Palestro CJ. 111 Indium pentetreotide scan detection of familial paragangliomas. Laryngoscope. 1998;108:228– 231. 148. Telischi FF, Bustillo A, Whiteman ML, et al. Octreotide scintigraphy for the detection of paragangliomas. Otolaryngol Head Neck Surg. 2000;122:358–362. 149. Bustillo A, Telischi FF. Octreotide scintigraphy in the detection of recurrent paragangliomas. Otolaryngol Head Neck Surg. 2004;130:479–482.
555
150. Litle VR, Reilly LM, Ramos TK. Preoperative embolization of carotid body tumors: When is it appropriate? Ann Vasc Surg. 1996;10:464–468. 151. Mitchell RO, Richardson JD, Lambert GE. Characteristics, surgical management, and outcome in 17 carotid body tumors. Am Surg. 1996;62:1034–1037. 152. Witterick IJ, Gullane PJ, Keller MA. Postoperative carotid body tumor evaluation: Analysis using MR angiography. Laryngoscope. 1995;105:764–767. 153. Shibuya Y, Umeda M, Yoshikawa T, et al. Carotid body tumour. Oral Oncol. 2002;38:313–317. 154. Smith JJ, Passman MA, Dattilo JB, et al. Carotid body tumor resection: Does the need for vascular reconstruction worsen outcome? Ann Vasc Surg. 2006;20:435–439. 155. Dickinson PH, Griffin SM, Guy AJ, et al. Carotid body tumour: 30 years experience. Br J Surg. 1986;73:14–16. 156. Hallett JW, Jr., Nora JD, Hollier LH, et al. Trends in neurovascular complications of surgical management for carotid body and cervical paragangliomas: A fifty-year experience with 153 tumors. J Vasc Surg. 1988;7:284–291. 157. de Vries EJ, Sekhar LN, Horton JA, et al. A new method to predict safe resection of the internal carotid artery. Laryngoscope. 1990;100:85–88. 158. Pantanowitz D, Davidge-Pitts K, Demetriades D. The significance of the carotid bifurcation angle in carotid body tumours. S Afr Med J. 1991;80:318–321. 159. Anand VK, Alemar GO, Sanders TS. Management of the internal carotid artery during carotid body tumor surgery. Laryngoscope. 1995;105:231–235. 160. Thabet MH, Kotob H. Cervical paragangliomas: Diagnosis, management and complications. J Laryngol Otol. 2001;115:467– 474. 161. Rosen IB, Palmer JA, Goldberg M, et al. Vascular problems associated with carotid body tumors. Am J Surg. 1981;142:459– 463. 162. Williams MD, Phillips MJ, Nelson WR, et al. Carotid body tumor. Arch Surg. 1992;127:963–967; discussion 967–968. 163. Maturo S, Brennan J. Baroreflex failure: A rare complication of carotid paraganglioma surgery. Laryngoscope. 2006;116:829– 830. 164. Timmers HJ, Karemaker JM, Wieling W, et al. Baroreflex control of muscle sympathetic nerve activity after carotid body tumor resection. Hypertension. 2003;42:143–149. 165. De Toma G, Nicolanti V, Plocco M, et al. Baroreflex failure syndrome after bilateral excision of carotid body tumors: An underestimated problem. J Vasc Surg. 2000;31:806–810. 166. Robertson D, Beck C, Gary T, et al. Classification of autonomic disorders. Int Angiol. 1993;12:93–102. 167. Smit AA, Timmers HJ, Wieling W, et al. Long-term effects of carotid sinus denervation on arterial blood pressure in humans. Circulation. 2002;105:1329–1335. 168. Robertson D, Hollister AS, Biaggioni I, et al. The diagnosis and treatment of baroreflex failure. N Engl J Med. 1993;329:1449– 1455. 169. Sharabi YY, Dendi RR, Holmes CC, et al. Baroreflex failure as a late sequela of neck irradiation. Hypertension. 2003;42:110– 116. 170. Timmers HHJ, Karemaker JJM, Lenders JJW, et al. Baroreflex failure following radiation therapy for nasopharyngeal carcinoma. Clin Auton Res. 1999;9:317–324. 171. Urquhart AC, Johnson JT, Myers EN, et al. Glomus vagale: Paraganglioma of the vagus nerve. Laryngoscope. 1994;104:440–445. 172. Biller HF, Lawson W, Som P, et al. Glomus vagale tumors. Ann Otol Rhinol Laryngol. 1989;98:21–26. 173. Durvasula VS, De R, Baguley DM, Moffat DA. Laser excision of glomus tympanicum tumours: Long-term results. Eur Arch Oto-rhino-laryngol. 2005;262:325–327. 174. Molony NC, Salto-Tellez M, Grant WE. KTP laser assisted excision of glomus tympanicum. J Laryngol Otol. 1998;112:956–958. 175. Robinson PJ, Grant HR, Bown SG. NdYAG laser treatment of a glomus tympanicum tumour. J Laryngol Otol. 1993;107:236– 237.
556
Goldstein et al.
176. O’Leary MJ, Shelton C, Giddings NA, et al. Glomus tympanicum tumors: A clinical perspective. Laryngoscope. 1991;101:1038–1043. 177. Jackson CG, Welling DB, Chironis P, et al. Glomus tympanicum tumors: Contemporary concepts in conservation surgery. Laryngoscope. 1989;99:875–884. 178. Gottfried ON, Liu JK, Couldwell WT. Comparison of radiosurgery and conventional surgery for the treatment of glomus jugulare tumors. Neurosurg Focus. 2004;17:E4. 179. Jackson CG, McGrew BM, Forest JA, et al. Lateral skull base surgery for glomus tumors: Long-term control. Otol Neurotol. 2001;22:377–382. 180. Murphy TP, Brackmann DE. Effects of preoperative embolization on glomus jugulare tumors. Laryngoscope. 1989;99:1244– 1247. 181. Briner HR, Linder TE, Pauw B, et al. Long-term results of surgery for temporal bone paragangliomas. Laryngoscope. 1999;109:577–583. 182. Manolidis S, Jackson CG, Von Doersten PG, et al. Lateral skull base surgery: The otology group experience. Skull Base Surg. 1997;7:129–137. 183. Green JD, Jr., Brackmann DE, Nguyen CD, et al. Surgical management of previously untreated glomus jugulare tumors. Laryngoscope. 1994;104:917–921. 184. Whitfield PC, Grey P, Hardy DG, et al. The surgical management of patients with glomus tumours of the skull base. Br J Neurosurg. 1996;10:343–350. 185. Tran BA, Huy P, Chao PZ, et al. Long-term oncological results in 47 cases of jugular paraganglioma surgery with special emphasis on the facial nerve issue. J Laryngol Otol. 2001;115:981–987. 186. Selesnick SH, Abraham MT, Carew JF. Rerouting of the intratemporal facial nerve: An analysis of the literature. Am J Otol. 1996;17:793–805; discussion 806–809. 187. Cece JA, Lawson W, Biller HF, et al. Complications in the management of large glomus jugulare tumors. Laryngoscope. 1987;97:152–157. 188. Spector GJ, Compagno J, Perez CA, et al. Glomus jugulare tumors: Effects of radiotherapy. Cancer. 1975;35:1316–1321. 189. Suit HD, Gallager HS. Intact tumor cells in irradiated tissue. Arch Pathol. 1964;78:648–651. 190. Jordan JA, Roland PS, McManus C, et al. Stereotastic radiosurgery for glomus jugulare tumors. Laryngoscope. 2000;110:35–38. 191. Kim JA, Elkon D, Lim ML, et al. Optimum dose of radiotherapy for chemodectomas of the middle ear. Int J Radiat Oncol Biol Phys. 1980;6:815–819. 192. Evenson LJ, Mendenhall WM, Parsons JT, et al. Radiotherapy in the management of chemodectomas of the carotid body and glomus vagale. Head Neck. 1998;20:609–613. 193. Mendenhall WM, Parsons JT, Stringer SP, et al. Radiotherapy in the management of temporal bone chemodectoma. Skull Base Surg. 1995;5:83–91. 194. Schild SE, Foote RL, Buskirk SJ, et al. Results of radiotherapy for chemodectomas. Mayo Clin Proc. 1992;67:537–540. 195. Springate SC, Weichselbaum RR. Radiation or surgery for chemodectoma of the temporal bone: A review of local control and complications. Head Neck. 1990;12:303–307. 196. Lybeert ML, van Andel JG, Eijkenboom WM, et al. Radiotherapy of paragangliomas. Clin Otolaryngol Allied Sci. 1984;9:105–109. 197. Powell S, Peters N, Harmer C. Chemodectoma of the head and neck: Results of treatment in 84 patients. Int J Radiat Oncol Biol Phys. 1992;22:919–924.
198. Valdagni R, Amichetti M. Radiation therapy of carotid body tumors. Am J Clin Oncol. 1990;13:45–48. 199. Mendenhall WM, Million RR, Parsons JT, et al. Chemodectoma of the carotid body and ganglion nodosum treated with radiation therapy. Int J Radiat Oncol Biol Phys. 1986;12:2175– 2178. 200. Mitchell DC, Clyne CA. Chemodectomas of the neck: The response to radiotherapy. Br J Surg. 1985;72:903–905. 201. Pemberton LS, Swindell R, Sykes AJ. Radical radiotherapy alone for glomus jugulare and tympanicum tumours. Oncol Rep. 2005;14:1631–1633. 202. Cole JM. Glomus jugulare tumor. Laryngoscope. 1977;87:1244– 1258. 203. Simko TG, Griffin TW, Gerdes AJ, et al. The role of radiation therapy in the treatment of glomus jugulare tumors. Cancer. 1978;42:104–106. 204. Cole JM, Beiler D. Long-term results of treatment for glomus jugulare and glomus vagale tumors with radiotherapy. Laryngoscope. 1994;104:1461–1465. 205. Foote RL, Coffey RJ, Gorman DA, et al. Stereotactic radiosurgery for glomus jugulare tumors: A preliminary report. Int J Radiat Oncol Biol Phys. 1997;38:491–495. 206. Liscak R, Vladyka V, Simonova G, et al. Leksell gamma knife radiosurgery of the tumor glomus jugulare and tympanicum. Stereotact Funct Neurosurg. 1998;70(Suppl 1):152–160. 207. Liscak R, Vladyka V, Wowra B, et al. Gamma knife radiosurgery of the glomus jugulare tumour—early multicentre experience. Acta Neurochir (Wien). 1999;141:1141–1146. 208. Eustacchio S, Trummer M, Unger F, et al. The role of Gamma knife radiosurgery in the management of glomus jugular tumours. Acta Neurochir Suppl. 2002;84:91–97. 209. Foote RL, Pollock BE, Gorman DA, et al. Glomus jugulare tumor: Tumor control and complications after stereotactic radiosurgery. Head Neck. 2002;24:332–338; discussion 338–339. 210. Lim M, Gibbs IC, Adler JR, et al. The efficacy of linear accelerator stereotactic radiosurgery in treating glomus jugulare tumors. Technol Cancer Res Treat. 2003;2(3):261–265. 211. Maarouf M, Voges J, Landwehr P, et al. Stereotactic linear accelerater-based radiosurgery for the treatment of patients with glomus jugulare tumors. Cancer. 2003;97:1093–1098. 212. Sheehan J, Kondziolka D, Flickinger J, et al. Gamma knife surgery for glomus jugulare tumors: An intermediate report on efficacy and safety. J Neurosurg. 2005;102(Suppl):241–246. 213. Krych AJ, Foote RL, Brown PD, et al. Long-term results of irradiation for paraganglioma. Int J Radiat Oncol Biol Phys. 2006;65:1063–1066. 214. Mendenhall WM, Amdur RJ, Hinerman RW, et al. Radiotherapy and radiosurgery for skull base tumors. Otolaryngol Clin North Am. 2001;34:1065–1077, viii. 215. Konefal JB, Pilepich MV, Spector GJ, et al. Radiation therapy in the treatment of chemodectomas. Laryngoscope. 1987;97:1331– 1335. 216. Verniers DA, Keus RB, Schouwenburg PF, et al. Radiation therapy, an important mode of treatment for head and neck chemodectomas. Eur J Cancer. 1992;28A:1028–1033. 217. Lalwani AK, Jackler RK, Gutin PH. Lethal fibrosarcoma complicating radiation therapy for benign glomus jugulare tumor. Am J Otol. 1993;14:398–402. 218. Elshaikh MA, Mahmoud-Ahmed AS, Kinney SE, et al. Recurrent head-and-neck chemodectomas: A comparison of surgical and radiotherapeutic results. Int J Radiat Oncol Biol Phys. 2002;52:953–956.
39 Pituitary Adenomas Mark Hornyak and William T. Couldwell
hormone. Although tumor cells may stain positively for a particular hormone on immunohistochemical analysis, they may not secrete functional hormone. Thus, a somatotrophic tumor may not cause acromegaly. Tumors that do not cause a clinical syndrome are called either silent or nonfunctional adenomas (3). Tumors containing cells that do not stain positively for any adenohypophyseal hormones and do not produce any clinical or biochemical evidence of secretion are termed nullcell tumors. The distribution of endocrinological tumor types can be found in Table 1. Large tumors can grow out of the sella turcica and cause pressure on delicate nervous structures, most commonly the optic chiasm, causing the classically described bitemporal hemianopsia. Microadenomas are defined as measuring less than 1 cm in their greatest dimension, whereas macroadenomas are greater than or equal to 1 cm. Some authors have augmented the classification with the term “mesoadenoma (4),” which has been used to describe adenomas measuring between 1 cm and 2 cm, reserving the term “macroadenoma” for even larger tumors. Most patients who are evaluated by a neurosurgeon harbor tumors larger than 1 cm, but the size of the tumor is often dependent on the endocrinological classification (Table 2). Microscopically, pituitary adenomas appear as monotonous sheets of cells with rounded nuclei and indistinct cytoplasmic borders (5). They are usually without conspicuous architectural features, but the monotony is often interrupted by delicate septae of connective tissue and by small perivascular spaces. Typically, the rounded nuclei contain delicate chromatin and small, peppery nucleoli (Fig. 1). Multinucleation and pleomorphism are not uncommon, but mitotic figures are unusual and frank anaplasia is rare (5). Tumors with significant anaplasia are often called atypical adenomas. The bland adenoma may be difficult to differentiate from normal pituitary gland, because cytology may be similar. The normal gland has a distinct acinar pattern, sinusoidal vascular spaces, and a mixed cellularity of acidophilic, basophilic, and chromophobic cell types (Fig. 2). The mixed population may be more easily appreciated on Orange G staining, and the architecture of the acini is better seen with a reticulin stain [Fig. 2(B)]. These features of normal pituitary tissue contrast with the larger, and more irregular, lobules of the pituitary adenoma that typically contain one cell type. Historically, tumors were classified on the basis of cytochromic staining characteristics with hematoxylin and eosin (H&E) stain—acidophilic adenomas were assumed to produce GH, basophilic adenomas were thought to secrete ACTH, and chromophobic adenomas were considered endocrinologically inactive (3). Currently, immunohistochemical staining is employed to determine the hormonal content of adenoma cells. Thus, these tumors are identified as producing one, or sometimes more (6), of the adenohypophyseal
INTRODUCTION The pituitary gland usually measures <1 cm3 in volume but, because of its importance in endocrinological homeostasis, it has been called the “master gland.” Because of the location of the pituitary gland at the base of the brain, surrounded by the bone of the skull base and critical neurovascular structures, the treatment of pituitary disease has been, and remains, a challenge.
INCIDENCE AND EPIDEMIOLOGY More than 4000 pituitary tumors were reported in the 2005– 2006 report by the Central Brain Tumor Registry of the United States (CBTRUS), making them a common type of brain tumor (1). They comprise up to 10% (6.3% in 2005–2006 CBTRUS) of all reported brain tumors and occur at an incidence rate of nearly 1 per 100,000 person-years (1). Although reports vary widely, a systematic review of the prevalence of pituitary tumors from pathological and radiographic studies provides an overall estimated prevalence of 16.7% (2). There is a slight female predominance, with a male/female ratio of 1:1.2. Pituitary tumors are diagnosed in all age groups, including the pediatric population, but the median age at diagnosis is 48 years, and these tumors are most common in the 65- to 74-year-old age group (1).
PATHOLOGY The most common tumor of the pituitary gland is the pituitary adenoma, which is typically a benign lesion arising from the secretory cells of the anterior pituitary gland. Because the pituitary is a functional gland and is lodged in a relatively tight anatomical space, tumors in this region can become clinically apparent either by hypersecretion or by mass effect on adjacent structures. Thus, pituitary tumors are most often classified by the hormone produced and by their size. More detailed classification schemes differentiate tumors on the basis of clinical and histological features as well. Hypersecretory syndromes include Cushing disease, hyperprolactinemia, acromegaly, hypergonadism, and secondary hyperthyroidism. These syndromes can occur with a tumor of any size but are often caused by small tumors. Typical antibodies used to recognize pituitary hormones include those to adrenocorticotropic hormone (ACTH), prolactin (PRL), growth hormone (GH), follicle-stimulating hormone (FSH), luteinizing hormone (LH), thyroid-stimulating hormone (TSH), and the α-subunit of the glycoprotein hormones (LH, FSH, and TSH). Tumors may contain populations of cells that stain positively for more than one pituitary 557
558
Hornyak and Couldwell
Table 1 Distribution of Pituitary Adenomas by Hormonal Production Hormone
Frequency
Prolactin Null cell tumor Growth hormone Adrenocorticotropic hormone Silent follicle-stimulating hormone/leuteinizing hormone Prolactin–growth hormone Silent adrenocorticotropic hormone Thyroid-stimulating hormone
30% 20% 20% 10% 10% 5% 2% 1%
hormones or hormonal components (i.e., α-subunit) (Fig. 3). Further immunostaining may reveal the presence of features specific for a particular tumor type. For instance, staining with CAM5.2 (low-molecular-weight keratin) in GH-positive cells may demonstrate fibrous bodies, which may have clinical implications [Fig. 3(C)] (7). On further analysis with routine immunohistochemical stains, pituitary adenomas typically stain positively with synaptophysin and neuron-specific enolase, and they do not stain with glial fibrillary acidic protein or cytokeratin. Null-cell adenomas may be identified with chromogranin A staining (8). Electron microscopy is also beneficial in demonstrating the ultrastructural features of cellular organelles and secretory granules. Despite these means of describing pituitary tumors, tissue architecture and cytoplasmic features do not always correlate well with the clinical aggressiveness of the lesion. Clinically and radiographically, tumors have been classified as the typical pituitary adenoma, invasive adenoma, atypical or anaplastic adenoma, or pituitary carcinoma. Invasive adenomas typically can be diagnosed preoperatively on magnetic resonance (MR) imaging or at the time of surgery. These tumors are found to spread outside the sella turcica and invade adjacent structures, most often the cavernous sinus, the clivus, or the sphenoid sinus. They comprise approximately 35% of tumors, occurring more commonly in macroadenomas and endocrinologically active tumors (Table 2) (9). The atypical adenoma will have the histological features of hypercellularity, increased mitotic figures, loss of tissue architecture, hyperchromatic nuclei, and frank anaplasia (Fig. 4). Pituitary carcinomas are rare tumors of the pituitary gland, with only approximately 140 cases reported in the English literature, comprising only 0.1% to 0.2% of all pituitary tumors (10). The hallmark of a pituitary carcinoma is distant metastases, regardless of cellular atypia. Sites of pituitary tumor metastasis include the brain, spinal cord, leptomeninges, bone, liver, lymph nodes, ovary, heart, and lung (11). The overall latency period between the presentation of a sellar adenoma and the manifestation of metastasis ranges from a few months to 18 years, with a median of 5 years. The majority of reported pituitary carcinomas (88%) Table 2
Pathological Features of Pituitary Adenomas by Tumor Type
Tumor type
Macro/Micro
Invasive
Prolactin adenoma Growth hormone Prolactin–growth hormone Adrenocorticotropic hormone Follicle-stimulating hormone/leuteinizing hormone Null cell tumor Thyroid-stimulating hormone
65% macro 85% macro 72% macro 15% macro 100% macro
52% 50% 31% 20% 21%
98% macro All macro
42% 75%
Figure 1 Histology of typical pituitary adenoma. (A) Photomicrograph (40x) demonstrating the histological features of a typical adenoma with H&E stain. A monotonous sheet of bland cells with round nuclei is shown. (B) Photomicrograph (40x, silver reticulin stain) demonstrating the lack of glandular architecture but some delicate fibrous septae.
are endocrinologically active, while null-cell pituitary carcinomas represent the remaining 12% of the reported cases (10). The histological and cytological characteristics of invasive adenomas and pituitary carcinomas vary from bland and monotonous, similar to typical adenomas, to grossly malignant. The immunohistochemical markers and vascular density of invasive adenomas and pituitary carcinomas are also similar to those of benign pituitary adenomas. Fifty percent of primary carcinoma tumors and most metastases display nuclear pleomorphism and/or hyperchromasia. Increased mitosis is also seen in pituitary carcinomas, but cellular atypia is not necessary for a diagnosis of carcinoma nor does it exclude benign adenoma. Signs of frank malignancy should raise concern for metastatic disease to the pituitary gland. Dural invasion by tumor cells may be an important histological sign of more aggressive tumors. In one large series, 46% of patients had microscopic evidence of dural invasion (12). The rate of dural invasion was higher in patients undergoing a second operation, in those with larger tumors, and in those with clinically nonfunctioning tumors. When these patients were monitored, however, tumor recurrence rate was
Chapter 39: Pituitary Adenomas
559
(A)
(B)
Figure 2 Histology of the normal pituitary gland. (A) Photomicrograph (40x, H&E) demonstrating the histological features of pituitary tissue including an acinar pattern with a variety of cell types. (B) Photomicrograph (40x, silver reticulin stain) demonstrating the clear acinar pattern of pituitary tissue. Source: Compare with Fig. 1.
not related to dural invasion in a consistent or significant fashion. Stains for mitotic activity (MIB-1, Ki-67) do correlate with tumor behavior. The MIB-1 monoclonal antibody, which recognizes the Ki-67 cell cycle-specific nuclear antigen, is a marker for cell division. In one study, establishing a threshold labeling index of 3% served to distinguish invasive from noninvasive adenomas with 97% specificity and 73% sensitivity and was associated with positive and negative predictive values of 96% and 80%, respectively (13). A clear and significant association is evident between p53 expression and tumor behavior, as the proportion of p53-positive cases among noninvasive adenomas, invasive adenomas, and pituitary carcinomas was 0%, 15.2%, and 100%, respectively. The mean Ki-67-derived growth fraction of p53-positive tumors is significantly higher than that of p53-negative tumors (14). Thus, these markers may have a role in differentiating aggressive tumors from those that act more benignly. Because of the variable methods of classifying pituitary tumors, the World Health Organization has proposed a fivetier classification system (15). Level 1 distinguishes tumors on the basis of clinical presentation and endocrine status (e.g.,
(C)
Figure 3 Pituitary adenoma immunohistology from a patient with acromegaly. (A) Photomicrograph (40x, GH immunoperoxidase stain) demonstrating a positive staining pattern for growth hormone. (B) Photomicrograph (40x, PRL immunoperoxidase stain) demonstrating a negative staining pattern for prolactin. (C) Photomicrograph (40x, CAM 5.2 immunologic stain) demonstrating fibrous bodies within adenoma cells.
560
Hornyak and Couldwell
Figure 4 Histology of anaplastic pituitary tumors. (A) Photomicrograph (40x, H&E) demonstrating atypical pituitary adenoma histology. The neoplastic cells exhibit moderately pleomorphic nuclei and small amounts of cytoplasm. There are occasional mitotic figures, and MIB-1 proliferation index was 11%. (B) Contrast-enhanced coronal T1-weighted MR image from patient shown in A demonstrating routine appearance of pituitary macroadenoma despite atypical histology. (C) Photomicrograph (40x, H&E) demonstrating malignant-appearing pituitary tumor. The neoplastic cells show severely pleomorphic nuclei with little cytoplasm. (D) Contrast-enhanced coronal T1weighted MR image from the patient shown in C demonstrating the invasive appearance of this aggressive lesion.
acromegaly). Level 2 is derived from preoperative imaging and intraoperative findings (e.g., macroadenoma, invasive). Level 3 is based on light microscopic investigations using routine staining methods (typically H&E) (e.g., typical vs. atypical/anaplastic). Level 4 classifies the tumor on the basis of immunohistochemical methods to establish the hormone or hormones produced by the tumor cells. Finally, level 5 is based on the ultrastructural features of the tumor cells (3). Thus, when this system is used, each tumor is graded on all five levels in a standardized fashion. The pathogenesis of pituitary adenomas, like that of most tumors, has not been well elucidated. This may be in part due to a lack of functional human pituitary tumor cell lines and appropriate animal models (16). Despite these difficulties, many factors have been implicated in the transformation of a normal adenohypophysial cell into an adenoma. The
normal hypothalamic releasing factors may have a role in tumorigenesis. Excessive hypothalamic hormone production or loss of negative feedback inhibition by pituitary cells may have a role in activating clonal expansion. Overexpression of growth factors and their receptors, specifically epidermal growth factor (EGF), transforming growth factor-α, EGF receptor, and vascular endothelial growth factor, has been identified in pituitary adenomas (17). A relationship between angiogenesis and tumor size, tumor invasiveness, and aggressiveness has been shown in some pituitary tumor types but not in others (18). Numerous oncogenes have been found to be activated in pituitary adenomas. The most abundant appear to involve gsp, cyclin D, and pituitary tumor transforming gene (PTTG) (16,19–21). The gsp oncogene is predominantly found in GH-producing tumors (40% of GHsecreting tumors in Caucasian patients). Mutation of the Gs protein constitutively activates agonist-activated receptors, including the GH receptor. Increased activity likely stimulates proliferation and hormone secretion (16). Cyclin D overexpression, which has been associated with growth control disturbances, may have a role in approximately 25% of pituitary tumors (16). PTTG, which may be the most important oncogene identified in pituitary tumors, has a role in cell proliferation. Low levels are required for proliferation, whereas high levels inhibit proliferation. Expression of PTTG has been found in 85% to 89% of pituitary tumors, but it is not expressed in normal pituitary tissue (22–23). Additionally, PTTG expression has been correlated with tumor invasiveness (22). Tumor suppressors may also have a role in the formation of pituitary tumors. Several familial types of adenomas that implicate tumor suppressor loss have been described. These include multiple endocrine neoplasia-type 1, Carney complex, isolated familial somatotropinoma, and von Hippel-Lindau syndrome (24). In sporadic pituitary adenomas, loss of retinoblastoma, p53, p16, and p27 tumor suppressor activity may play a role in neoplasia (19–21,24). Although pituitary adenomas constitute the vast majority of tumors within the pituitary gland, other tumors include the granular cell tumor and the pituicytoma. These tumors are very rare and are found within the neurohypophysis or the infundibulum. Other pathological entities involving the pituitary gland include lymphocytic hypophysitis, giant cell hypophysitis, and empty sella syndrome. Tumors that can be found within the sella turcica and in the perisellar areas include the craniopharyngioma, germinoma, chordoma, astrocytoma (usually of the optic chiasm or hypothalamus), esthesioneuroblastoma, meningioma, fibrosarcoma, metastatic disease, Langerhans histiocytosis, chondroma, giant cell tumors and other neoplasms of the calvarium, gangliocytoma, and schwannoma. Cystic lesions occurring in the area of the sella include Rathke cleft cysts, epidermoid cysts, and arachnoid cysts. Additional entities include hypophysitis, sarcoidosis, mucocele, abscess, and even aneurysm. Table 3 lists several pathological processes that may affect the area of the sella turcica. Although the clinical picture will provide information regarding a patient’s diagnosis, these entities should be considered, especially in the patient with an atypical presentation.
STAGING The typical tumor staging system based on tumor size, lymph node involvement, and distant metastases is not applicable to tumors of the pituitary gland.
Chapter 39: Pituitary Adenomas Table 3
Lesions of the Sella Turcica
Primary tumors of the sella Pituitary neoplasia (adenoma, carcinoma) Meningioma Craniopharyngioma Glial tumor (optic or hypothalamic glioma) Pituicytoma (infundibuloma) and granular cell tumor of the infundibulum Germ cell tumor Langerhans cell histiocytosis (Hand–Sch¨uller–Christian disease) Chordoma Hypothalamic hamartoma Choristoma Gangliocytoma Tumors of bone (chondrosarcoma) Esthesioneuroblastoma Schwannoma Secondary tumors Metastatic carcinoma Hematopoietic neoplasm (lymphoma, leukemia) Cysts Rathke cleft cyst Epidermoid cyst Arachnoid cyst Mucocele Inflammatory lesions Lymphocytic hypophysitis Granulomatous hypophysitis Sarcoidosis Infection Bacterial abscess Tuberculoma Other non-neoplastic lesions Pituitary hyperplasia Empty sella syndrome Pituitary apoplexy (usually tumor-related) Aneurysm
TREATMENT A multidisciplinary approach is important when evaluating and treating a patient with pituitary disease. Cooperation is needed from the patient’s primary care physician as well
561
as an endocrinologist before surgical intervention is undertaken, and consultation by an ophthalmologist and radiation oncologist often can be beneficial. The primary treatment for most pituitary tumors is surgery; however, a complete evaluation is necessary before intervention. Certain tumors (e.g., prolactinomas) can be treated medically, while some invasive tumors will require radiation.
History and Physical Examination The initial step in evaluating the patient with pituitary disease is to obtain an account of the illness. A complete medical history with a thorough endocrinological history is essential. The historian should elicit information about symptoms of an endocrinological disorder, visual disturbance, and headaches. Common features of endocrinopathy include decreased libido and erectile dysfunction in men, irregular menses or amenorrhea and galactorrhea in reproductive-aged women, and fatigue or general weakness. Signs of acromegaly include enlargement of the hands, feet, and facial bones; joint pain; skin tags; and hyperhydrosis (Fig. 5). Cushing syndrome usually causes weight gain, fatigue and weakness, insomnia, irritability, cognitive and mood changes, and osteoporosis with bone fractures. New onset or worsening of previously diagnosed hypertension or diabetes may also be seen. The patient with hypercortisolemia often presents with the classically described moon facies, facial plethora, central adiposity, buffalo hump, and abdominal striae [Fig. 6(A,B)]. These changes can resolve with appropriate treatment [Fig. 6(C,D)]. The historian should also elicit from the patient whether there is a history of excessive thirst or fluid intake, which is a sign of diabetes insipidus (DI). Pituitary adenomas only very rarely will cause DI, and its presence, which can be confirmed with urine and serum studies and a water deprivation test, generally implicates another type of sellar or suprasellar lesion. Physical examination may be unremarkable in many patients; however, the signs of advanced acromegaly, Cushing syndrome, and hyperprolactinemia (in women) are often easily recognized. Hypogonadism can also be evaluated in the male patient, and subtle signs of hyperthyroidism can also be found on physical examination. A thorough visual examination should be performed, and a patient with any visual complaint or finding should be referred for evaluation with visual field testing by an ophthalmologist.
Figure 5 Clinical appearance of acromegaly. The facial features (A) and hand changes (B) of acromegaly are demonstrated in this 65-year-old woman, who harbored a GH-secreting pituitary tumor.
562
Hornyak and Couldwell
Figure 6 Clinical appearance of Cushing syndrome. The facial features (A) and abdominal changes (B) of hypercortisolemia are demonstrated in this 40-year-old woman with an ACTH-secreting pituitary tumor. The facial (C) and abdominal changes (D) have resolved 32 months after surgical resection of the pituitary tumor.
Laboratory Tests A complete endocrinological panel of laboratory studies should also be performed preoperatively (Table 4). These evaluations will help in the diagnosis of any hypersecretory syndrome but can also detect a hypopituitary state. At a minimum, the serum concentrations of the following hormones should be ascertained: PRL, LH, FSH, testosterone (men), estrogen (women), insulin-like growth factor-1 (IGF-1, or somatomedin-C), GH, ACTH, cortisol, TSH, and free T4. Because of variations in plasma cortisol levels throughout the day, levels should be drawn in the morning before eating. Similarly, because of the pulsatile release of growth hormone, Table 4 Hormones to be Tested in Laboratory Investigations for Evaluation of the Patient with a Pituitary Adenoma or Other Sellar Lesion Prolactin (PRL) Thyroid-stimulating hormone (TSH) Free T4 Growth hormone (GH) (fasting, A.M.; consider oral glucose tolerance test) Insulin-like growth factor (IGF-1, Somatomedin-C) Adrenocorticotropic hormone (ACTH) Cortisol (fasting, A.M.) Leuteinizing hormone (LH) Follicle-stimulating hormone (FSH) Estrogen (E2) (women) Testosterone (men)
its level should also be tested in the morning while fasting. If GH levels are inconclusive, an oral glucose tolerance test should be performed, and GH levels should suppress to less than 1 µg/L after a glucose load (25). At the same time, a metabolic panel to evaluate for signs of DI and a routine complete blood count can be performed as well. These laboratory investigations are generally adequate to diagnose all pituitary endocrinopathy except Cushing disease. The differentiation between the different causes of hypercortisolemia may require low- and high-dose dexamethasone suppression tests, a corticotropin-releasing hormone stimulation test, and possibly inferior petrosal sinus sampling (26).
Radiology If signs of a hypersecretory syndrome or symptoms of a mass lesion indicate that a pituitary lesion may exist, a radiographic evaluation must be undertaken. MR imaging is the study of choice, and it should include thin-cut (2 mm) sections through the sella turcica in the axial, sagittal, and coronal planes. The sella should be evaluated with pre- and post-contrast-enhanced T1-weighted images, as well as T2weighted images. The normal anterior lobe of the pituitary gland appears iso- to slightly hyperintense on T1 images. It is usually less than 1 cm in height, but, in times of increased endocrine activity (e.g., puberty, pregnancy), the gland can enlarge (27) and have a more intense signal on T1 images. The posterior lobe of the pituitary gland can also be visualized; it is usually small and very intense on T1-weighted images and
Chapter 39: Pituitary Adenomas
Figure 7 Normal pituitary gland on MR imaging. An unenhanced sagittal T1-weighted image demonstrates the normal appearance of the pituitary gland and the sella turcica. Note the size of the sella and pituitary gland, as well as the pituitary stalk and posterior pituitary gland (neurohypophysis), which is seen as the “bright spot” posteriorly within the sella.
has thus been called the “bright spot” (Fig. 7) (28). The normal infundibulum is often seen as well and can deviate from the midline in nearly half of normal studies (29). Pituitary macroadenomas are easily visualized. They are large masses in the region of the sella turcica in close proximity to the pituitary gland (Fig. 8). They commonly have suprasellar extension but can also invade the cavernous sinus, erode the clivus, or grow into the sphenoid sinus. Tumors with suprasellar extension may take on the characteristic “snowman” shape, in which they are constricted as they grow up through the diaphragma sellae. These large tumors are usually isointense to the normal pituitary gland on MR imaging. They enhance, but not as vividly or as early after gadolinium administration as the normal gland. Tumors may be cystic and appear heterogeneous. When evaluating MR studies of a macroadenoma, some important features to note are the extension of the tumor out of the sella turcica, the presence and amount
563
of optic chiasm compression, and extension of tumor into the cavernous sinus. If tumor extends beyond the suprasellar cistern, a routine trans-sphenoidal approach for surgery may be inadequate, and an extended trans-sphenoidal or transcranial approach may be needed. If there is radiographic evidence of optic compression, ophthalmological consultation should be sought, and the lesion should be treated urgently. If a tumor has invaded the cavernous sinus, evidenced by tumor lateral to the cavernous segment of the internal carotid artery (30,31), complete resection is unlikely (Fig. 9), and other treatments should be considered either primarily or as an adjunct to subtotal resection. Microadenomas, unlike the larger tumors, may be difficult to visualize on MR imaging (Fig. 10). Very small lesions are sometimes not visible with the resolution of the typical MR scanner, but they can produce hypersecretory syndromes. When the presence of a microadenoma is suspected but not visualized on routine MR imaging of the sella, other clues may disclose the location of the adenoma. For example, the pituitary stalk may deviate away from the side of the lesion, or the sella may be expanded on the side of the tumor. The bony contours of the sella are best evaluated using computed tomography (CT) imaging, and these images can also be useful in planning the operation, by demonstrating the bony architecture of the sphenoid sinus (32), and in computer-assisted frameless surgical navigation (33). If a sphenoid septum is present, it is important to note that the sella floor may bulge to the opposite side, mimicking changes seen with an adenoma. If a microadenoma is not visualized on routine MR imaging, a dynamic study that includes delayed images after gadolinium administration may be beneficial (34). Hemorrhage into a pituitary tumor (pituitary apoplexy) can be visualized on routine MR imaging as blood and blood products causing heterogeneity of the signal within the tumor (Fig. 11). Signal characteristics will depend on the age of the hematoma, but gradient echo T2-weighted images can be helpful in distinguishing a cystic tumor from apoplexy [Fig. 11(C)].
SURGERY In general, the goals of surgery for pituitary tumors are (i) to eliminate tumor mass, (ii) to normalize hormonal hypersecretion, (iii) to preserve normal pituitary function,
Figure 8 Pituitary macroadenoma on MR imaging. (A) Contrast-enhanced coronal T1-weighted image demonstrating a macroadenoma with suprasellar extension. Note the bowing of the medial walls of the cavernous sinuses and the deformation of the optic chiasm (arrow). (B) Contrast-enhanced sagittal T1-weighted image demonstrating a macroadenoma. Again the suprasellar extension and expansion of the sella into the sphenoid sinus can be noted.
564
Hornyak and Couldwell
Figure 9 Invasive adenoma on MR imaging. Preoperative unenhanced (A) and contrast-enhanced (B) coronal T1-weighted images demonstrating a pituitary macroadenoma with invasion into bilateral cavernous sinuses. (C) Postoperative contrast-enhanced coronal T1-weighted image demonstrating residual tumor within the cavernous sinuses.
(iv) to minimize potential for recurrence, and (v) to obtain a specimen for definitive tissue diagnosis and prognosis (35). The surgical indications for various tumors differ. For instance, pituitary apoplexy is an indication for emergent resection of a pituitary tumor (36–39). The sudden hemorrhage
Figure 10 Pituitary microadenoma on MR imaging. (A) Unenhanced coronal T1-weighted image of microadenoma, which is difficult to discern. It can be seen as a hypointensity within the left portion of the sella (arrow). (B) Contrast-enhanced coronal T1-weighted image demonstrates vivid enhancement of pituitary gland and less enhancement of tumor, making the microadenoma (arrow) more visible.
into a necrotic adenoma causes a rapid increase in tumor size, often causing compression of the optic chiasm or cranial nerves within the cavernous sinus. It is associated with a sudden onset of severe headache, visual loss, cranial neuropathy, and acute adrenal insufficiency. Rapid decompression of the optic nerves preserves visual function. Before surgical intervention, corticosteroid supplementation is necessary. With the exception of prolactinomas, surgery is indicated for all macroadenomas that cause clinical or radiographic compression of the optic apparatus. In the case of prolactinomas, surgery is indicated in the event that dopamine agonist therapy is ineffective in normalizing serum PRL levels or in reducing mass effect, or when patients fail to tolerate medical treatment. Surgical debulking of a large prolactinoma may reduce the dose of dopamine agonist medication required, and thus improve tolerability. Surgery can be considered as an alternative to medical therapy in patients with microprolactinomas who do not want to remain on longterm medical therapy. This should be considered, especially in young patients, as the economical impact of an uncomplicated operation is similar to that for 10 years of medical therapy (40). Surgery can also be considered in the case of large, cystic prolactinomas, because the cystic component is unlikely to shrink with medical treatment and mass effect will not be relieved. Microprolactinomas are much more likely to be cured surgically than are macroprolactinomas (41–44). Biochemical cure rates as high as 91% have been reported and may be higher in patients with low serum prolactin levels (less than 200 ng/mL) (41). Resection of a microprolactinoma should not be attempted unless the tumor can be removed completely and a biochemical cure is the expected outcome (45). . Resection is the primary treatment in other hormonesecreting microadenomas. Although medical therapy for acromegaly is effective, surgery remains the primary treatment for GH-secreting tumors. Resection results in a prompt decrease in GH levels (46–48). Long-term biochemical cure (sustained GH < 5 ng/mL) can be achieved in 61% to 87% of noninvasive microadenomas and 23% to 68% of larger, more invasive tumors (47,49–51). Even in invasive macroadenomas, debulking of the tumor increases the likelihood of achieving biochemical disease control with medical therapy (52,53). There is currently no effective pharmacological agent that can reduce ACTH production in Cushing disease. Surgery is therefore the treatment of choice, and long-term control can be achieved in 57% to 85% of surgically treated patients (54–58). Surgery is also the treatment of choice for other hormone-secreting (gonadotrophin, TSH) adenomas (59,60).
Chapter 39: Pituitary Adenomas
565
Figure 11 Pituitary apoplexy on MR imaging. (A) Unenhanced sagittal T1-weighted image demonstrating a macroadenoma. The MR imaging shows heterogeneous signal intensities including diffuse hyperintensity, which is consistent with subacute hemorrhage. (B) Contrast-enhanced coronal T1-weighted image demonstrating macroadenoma, again with signal heterogeneity consistent with hematoma. (C) Axial gradient echo T2-weighted image demonstrating T2 signal loss and blooming artifact within the sella diagnostic for blood products.
Large or symptomatic nonfunctioning adenomas are also treated with surgery. These tumors most often present with visual disturbance or hypopituitarism. In one report with 6 years of follow-up monitoring, surgery effectively controlled these tumors in approximately 90% of patients (61). The management strategy for incidentally discovered adenomas is somewhat controversial. Small pituitary adenomas may be found radiographically in up to 22% of the population (2). Although no consensus exists regarding treatment of patients with these “incidentalomas,” it is generally agreed that they should be evaluated with screening for endocrinological function (either hypersecretory state or hypopituitarism) and visual deficit. This strategy applies even if the patient has no complaint, especially if the tumor is a macroadenoma, because 34% of patients may have an unnoticed visual deficit and 41% may have an endocrinological problem (62–67). If the screening tests are normal, the patient must be observed with serial MR imaging, because tumors increase in size in 12% to 50% of patients monitored (63,64,68). In a recent report, 4 of 42 patients with macro-incidentalomas monitored for a mean of 62 months suffered an apoplexy during follow-up care (68). All four developed hypopituitarism and one has long-term visual disturbance secondary to the hemorrhage. Original tumor size, patient age, and the presence of intratumoral cysts were not correlated with tumor growth. Apoplexy was correlated with tumor growth but not with initial size of the lesion. Therefore, given the risk of apoplexy, surgery may be considered in young patients found to harbor a tumor, especially if the tumor is greater than 15 mm in size. Surgery should certainly be offered to any patient who exhibits growth of, or symptoms arising from, an incidentally discovered adenoma (69). When it has been decided that resection of a pituitary adenoma is indicated, the operation must be planned. Unless contraindicated, the trans-sphenoidal route is most commonly employed, but a craniotomy is sometimes necessary. The technical details of these surgical approaches are discussed elsewhere in this text. Regardless of the primary mode of therapy, patients with pituitary tumors should be evaluated by both an endocrinologist and the surgical team before treatment. In the case of elective surgery, the patient harboring disease of the pituitary gland should have his or her pituitary function eval-
uated before operative intervention. Cortisol should be replaced as needed, and stress doses of corticosteroids should be given perioperatively. Thyroid hormone should also be supplemented when necessary, and in the case of central hyperthyroidism, beta-blockers should be given for prophylaxis of cardiac arrhythmias (70). Postoperative care should include careful monitoring for neurological changes and signs of cortisol deficiency (Addisonian crisis) or DI. Corticosteroid supplementation should be given to patients with Cushing disease or known ACTH deficiency. Those patients with intact hypothalamic– pituitary–adrenal (HPA) axis (morning cortisol >16 µg/dL) do not need postoperative supplementation. If the integrity of this system is in doubt, perioperative stress–dose steroids followed by physiological replacement should be given until the HPA axis can be tested (71).
RADIATION Surgery for pituitary adenomas is safe and effective, and it remains the primary treatment of most pituitary adenomas; however, when tumors are large and invasive, surgery is not as effective. Recurrence rates for large, invasive, hormonally active tumors can be as high as 93% (72), and radical resection carries a risk of permanent cranial neuropathy. Radiation therapy has long been used to treat residual or recurrent pituitary adenomas. Conventional external beam radiation therapy (XRT) has had good results in terms of controlling tumor growth. This therapy can be administered either as an adjunct to resection or as a primary therapy. In either case, the use of XRT has growth control rates of 82% to 97% (73–83). The efficacy of XRT in normalizing hormone hypersecretion is considerably lower, with control rates of 38% to 83% (73,74,80,81,84). Typical treatment dose is 45 to 50 Gy divided over 25 to 30 fractions. Hypopituitarism is common after XRT, with approximately 50% of patients requiring hormone replacement 20 years after treatment (73). Some authors believe that all patients will experience hypopituitarism after pituitary irradiation given long enough follow-up. XRT carries a low risk of other complications, including cerebral radiation damage, cranial neuropathy including visual deterioration, cerebrovascular injury and stroke, and secondary malignancy (85).
566
Hornyak and Couldwell
Figure 12 SRS treatment plan. Axial (A), coronal (B), and sagittal (C) demonstrations of a treatment plan for radiosurgery to be delivered to the left cavernous sinus. The isocenter is denoted as a “+.” Isodose lines demonstrate the high dose (80%) delivered to the tumor (broad orange line), while the optic apparatus and pituitary gland are relatively spared (<20% of maximum dose). (This is the same patient as shown in Fig. 9.)
By definition, stereotactic radiosurgery (SRS) is a high dose of radiation delivered precisely, usually in a single session (86). SRS can reduce the dose of radiation delivered to the brain, pituitary gland, optic apparatus, and cranial nerves while providing an adequate dose to the tumor (Fig. 12). With the proven efficacy of SRS in various tumors, a greater emphasis has been placed on reducing the morbidity of surgery by limiting resection of benign intracranial lesions in favor of radiosurgical treatment of residual and recurrent disease (87,88). Pituitary surgery can be tailored to postoperative SRS by limiting resection to preserve cranial nerve and endocrine function. For instance, when tumor is left within the cavernous sinus and postoperative radiosurgery is anticipated, the pituitary gland can be transposed away from the radiosurgical target and held away by an interposing fat graft (89,90), thereby limiting the radiation dose to the functioning gland and reducing the risk of hypopituitarism. Pituitary tumors treated with SRS typically receive a marginal dose of radiation of approximately 16 Gy (reported range 12–28 Gy). The dose to the optic pathways should be limited to less than 12 Gy, and even below 8 Gy when possible. Thus, SRS is usually limited to treating tumors less than 3 cm in diameter and those with space between the lesion to be treated and the optic apparatus. The use of SRS can provide growth control rates of greater than 90%. Endocrinological outcomes are not as good, with successful cure of endocrinopathy ranging from 0% to 100% of all functionally secreting tumors and weighted average cure rates of approximately 50% for Cushing disease and acromegaly and 25% for prolactinomas (91,92). SRS also has risks similar to those of XRT, including hypopituitarism, cerebral radiation injury, visual loss, and secondary neoplasia. Because of the potential risks of radiation therapies, XRT and SRS should be reserved for treatment of residual or recurrent disease in those patients who have undergone resection of pituitary tumor, those who are not surgical candidates, and those who refuse surgery (78,93,94). Although conventional XRT has been used to treat pituitary tumors (93), it has essentially been replaced by conformal radiotherapy techniques. Fractionated stereotactic radiotherapy (95) and intensity-modulated radiotherapy (96) can be used to treat
larger tumors or those in close proximity to the optic pathways. Fractionated stereotactic treatment follows the same radiobiological principles of XRT but with greater precision of planning, and thus fewer expected side effects. Control rates are very high, reported at 98% to 100% (95,97–99), with normalization of hormone hypersecretion in 38% to 84% of patients (95,97,98). There is still a risk of hypopituitarism, cerebral injury, and visual loss, but these risks appear less than with conventional XRT.
CHEMOTHERAPY—MEDICAL THERAPY Prolactinomas are an exception to the preferred use of surgery as the primary treatment for pituitary tumors because medical therapy is very effective in their treatment. Medical treatments can also be used as adjuncts to surgical treatment for acromegaly and Cushing disease. The prototypical, and most common, medication used to treat prolactinoma is bromocriptine, a D2 receptor agonist. This drug is taken orally, at a usual dose of 5 to 10 mg daily, and can normalize serum prolactin levels in up to 90% of patients, with tumor shrinkage seen in approximately 85% (100–104). Bromocriptine is a safe treatment, but it has intolerable side effects in approximately 10% of patients and may also be required as a lifelong treatment in some patients (105). Additionally, up to 25% of tumors are resistant to bromocriptine (106,107). Cabergoline and pergolide, other dopamine agonists, have improved tolerability profiles but still may be ineffective in 10% to 15% of patients (106). Cabergoline is currently the drug of choice in treating prolactinomas because it is better tolerated than bromocriptine. This drug is usually given orally at a dose of 0.25 mg weekly, and it may be increased every 4 weeks as needed, up to 1 mg two times a week. Prolactinomas are very responsive to medical treatment, and recent literature indicates that normalization of PRL levels with cabergoline for several years may definitively treat the tumor, with no evidence of recurrence after discontinuation of medication (Fig. 13) (108). However, the long-term effectiveness of this strategy has not been defined (109). Other medications such as lisuride and quinagolide are being used but will likely yield similar efficacy, in the range of 90%. Surgery
Chapter 39: Pituitary Adenomas
567
Figure 13 Medical treatment of prolactinoma. Contrast-enhanced coronal T1-weighted images demonstrating resolution of a large prolactinoma with medical therapy. (A) Image obtained at the time of diagnosis. Serial follow-up imaging shows shrinkage of the tumor at 6 (B), 18 (C), and 36 (D) months after initiation of therapy.
is therefore indicated when medical therapy fails to correct hyperprolactinemia or is not tolerated by the patient. In the case of macroprolactinomas, surgical cure is achieved in less than 50% of patients (41,110); however, even if a cure cannot be achieved, surgery will decompress the optic apparatus (if compression is present) and cytoreduction will improve the remaining tumor’s responsiveness to dopamine agonist therapy, lowering the required dose, thus improving tolerability (111,112). One novel therapy for prolactinoma targets the tumor’s blood supply. A rodent model of prolactinomas has been used to test the antivascular agent ZD6126. It has demonstrated antitumoral effects in other tumor types (113) and may prove effective in treating prolactinomas and other pituitary tumors (114). Although acromegaly is primarily treated surgically, patients who refuse surgery or have recurrent or residual disease postoperatively can be treated medically. Normalization or reduction of GH and IGF-1 levels leads to improved survival (115). Somatostatin analogs are the main line of medical therapy, with octreotide being the prototypical drug. It is delivered subcutaneously at a typical dose of 100 to 250 µg three times daily; longer-acting analogs are also available. This class of drugs successfully reduces serum GH and IGF-1 levels in 50% to 70% of patients, with normalization
of IGF-1 in approximately 30% after failed pituitary surgery (116). A GH receptor antagonist has also been used effectively to treat acromegaly: Pegvisomant, when administered in a dose of 10 to 20 mg subcutaneously daily, can normalize IGF-1 in 80% to 90% of patients with acromegaly (117). Prolactin may be cosecreted with GH, or tumors may stain for PRL on immunohistochemical staining, in nearly half of patients with acromegaly. Thus, dopamine agonist therapy has been used with some success (118–120). The selective estrogen receptor modulator, raloxifene, has been used to treat refractory acromegaly. In a series of eight men with persistent acromegaly despite aggressive management, raloxifene reduced IGF-1 by a mean of 16% and normalized it in two patients (121). Recently, many subtypes of somatostatin receptors have been identified (122), and ligands to these receptors may have a role in the treatment of tumors secreting GH or GH and PRL in the future (123,124). Ketoconazole, metyrapone, mitotane, aminoglutethimide, and trilostane are steroidogenesis inhibitors that can be used to decrease serum cortisol levels in Cushing disease. Their effect and tolerability are limited (125,126), and therefore they should only be used for short-term treatment or when no other treatment is available. Adrenalectomy is also an option in refractory cases. It effectively treats
568
Hornyak and Couldwell
hypercortisolism, but the risk of Nelson syndrome should be considered, as it may occur in up to 80% of patients with Cushing syndrome after total adrenalectomy (127). An emerging medical treatment for refractory Cushing disease targets the peroxisome proliferator-activating receptor-γ (PPAR-γ ), which is abundantly expressed in normal and tumor cells that secrete ACTH (128). The thiazolidinediones are a group of PPAR-γ agonists that have suppressed pituitary tumor growth in animal models (128,129). Rosiglitazone, a member of this class of durgs usually administered for type-2 diabetes mellitus to improve insulin resistance, has been used with some success to treat Cushing disease (130–132). In one study of 14 patients with Cushing disease, rosiglitazone was able to normalize urinary-free cortisol in 6 patients, and this was maintained in 4 patients for 7 months of follow-up care; the remaining 8 patients had no response (130). Thiazolidinediones therefore may have a role in the treatment of some ACTH-producing tumors and possibly other pituitary tumors as well (133,134). For tumors secreting TSH, somatostatin analogs can be used when surgery fails (135), and antithyroid drugs (methimazole or propothiouracil) have been used to reduce thyroid hormone levels preoperatively (59,125). Typical cytotoxic chemotherapeutic agents have been used in small numbers of patients with aggressive pituitary adenomas and carcinomas that have failed to respond to the usual treatment paradigm of surgery, radiation, and biotherapy. Agents that have been used include CCNU with 5-FU, Gliadel wafers, and temozolomide (136–138). Although the CCNU/5-FU regimen showed little promise, five of eight patients treated with Gliadel wafers showed at least stabilization of the disease (137). Several patients with aggressive pituitary tumors, including both invasive adenomas and carcinomas, have been successfully treated with temozolomide (138,139,140).
OUTCOME AND PROGNOSIS In experienced hands, trans-sphenoidal surgery applied to pituitary adenomas is a low-risk operation. In recent reports of large series of patients, rates of permanent morbidity are less than 5%, with mortality rates well under 1% (46,47,141–144). This represents a modest but definite improvement from the pre-CT and pre-MR imaging era (145,146). Most postoperative complications can usually be anticipated and treated. The most common immediate complications are cerebrospinal fluid (CSF) rhinorrhea and DI. Both these complications are more common after removal of larger tumors, and thus the surgeon must be vigilant in observing for these conditions. Postoperative leak of CSF has been traditionally reported in approximately 4% of cases but occurs at a much lower rate in the senior author’s recent experience (147). A careful history and physical examination should be performed daily with provocative testing for CSF rhinorrhea. If a leak is found, the patient should be returned to the operating room for packing or repacking of the sella. After surgery for pituitary adenoma, DI occurs in approximately 16% of patients, requires treatment in only 11%, and persists in only approximately 0.5% (148). Careful monitoring of the patient’s urine output and specific gravity, as well as serum osmolality and sodium concentration, can lead to the diagnosis. In most cases DI is transient, lasting only 1 to 3 days postoperatively. DI may require vasopressin or desmopressin, but this is not routine after resection of a pituitary adenoma, and most patients can be treated with liberal free water intake. Other less common
complications occurring after trans-sphenoidal surgery include visual loss (149), meningitis (47,150), and vascular injury (151). In the long term, nasal complications may develop (152), and panhypopituitarism may be seen in up to 5% of patients (47,143). Surgery on pituitary adenomas successfully treats mass effect in nearly all cases (with gross total resection in 65%, limited by larger, invasive tumors) (153) and is acceptable treatment of endocrinopathy in most patients. After resection of a nonfunctioning adenoma, growth of residual or recurrent tumor is seen in approximately 10% of patients (61,154). Rates of successful endocrinological outcome vary among tumor type and size and length of follow-up. Reported endocrinologic cure rates after surgery range from 42% to 85% (40,41,46,47,51,56,72,110,142,143,153,155–159). Although most pituitary tumors can be treated effectively with a single mode of therapy (with long-term control rates of approximately 90% for typical tumors), patients with invasive tumors have a worse prognosis and often require multimodal therapy (160). Tumors that fail a single mode of therapy can often be treated successfully with another. Patients suffering from persistent or recurrent acromegaly after trans-sphenoidal surgery can often be successfully treated medically (116). Those tumors that do not respond to surgery and medical management can then be treated with radiotherapy concurrent to medical therapy. Similarly, radiation can be used postoperatively to reduce recurrence rates when medical therapy is unavailable or is not tolerated. In patients who present with regrowth of tumor that has been resected via a trans-sphenoidal operation, a second surgery is often warranted and is successful in most cases (161). Overall, multimodal therapy is successful in the vast majority of patients with pituitary ademomas (115,126,162). A poorer prognosis exists for patients with large, invasive tumors, and those with pituitary carcinoma. All patients must be followed carefully for both radiographic and endocrinological recurrence so that further treatment can be provided without delay. In the case of a multidisciplinary treatment team the outcome and prognosis for the great majority of patients in whom pituitary adenomas are diagnosed is good, and it will continue to improve as surgical techniques, medical therapies, and radiation delivery improve.
ACKNOWLEDGMENTS We thank Kristin Kraus for her editorial assistance in preparing this chapter. REFERENCES 1. CBTRUS. Statistical Report: Primary Brain Tumors in the United States, 1998–2002. Hinsdale, IL: Central Brain Tumor Registry of the United States. 2005. 2. Ezzat S, Asa SL, Couldwell WT, et al. The prevalence of pituitary adenomas: A systematic review. Cancer. 2004;101:613–619. 3. Kovacs K, Horvath E, Vidal S. Classification of pituitary adenomas. J Neurooncol. 2001;54:121–127. 4. Scamoni C, Balzarini C, Crivelli G, et al. Treatment and longterm follow-up results of prolactin secreting pituitary adenomas. J Neurosurg Sci. 1991;35:9–16. 5. Burger PC, Scheithauer BW, Vogel FS. The region of the sella turcica. In: Surgical Pathology of the Nervous System and Its Coverings, 4th ed. Philadelphia, PA: Churchill Livingstone, 2002:437–498. 6. Scheithauer BW, Horvath E, Kovacs K, et al. Plurihormonal pituitary adenomas. Semin Diagn Pathol. 1986;3:69–82.
Chapter 39: Pituitary Adenomas 7. Nishioka H, Haraoka J, Akada K. Fibrous bodies are associated with lower GH production and decreased expression of Ecadherin in GH-producing pituitary adenomas. Clin Endocrinol (Oxf). 2003;59:768–772. 8. Lloyd RV, Scheithauer BW, Kovacs K, et al. The immunophenotype of pituitary adenomas. Endocr Pathol. 1996;7:145– 150. 9. Scheithauer BW, Kovacs KT, Laws ER Jr, et al. Pathology of invasive pituitary tumors with special reference to functional classification. J Neurosurg. 1986;65:733–744. 10. Ragel BT, Couldwell WT. Pituitary carcinoma: A review of the literature. Neurosurg Focus. 2004;16 (4):E7. 11. Scheithauer BW, Kurtkaya-Yapicier O, Kovacs KT, et al. Pituitary carcinoma: A clinicopathological review. Neurosurgery. 2005;56:1066–1074; discussion 1066–1074. 12. Meij BP, Lopes MB, Ellegala DB, et al. The long-term significance of microscopic dural invasion in 354 patients with pituitary adenomas treated with transsphenoidal surgery. J Neurosurg. 2002;96:195–208. 13. Thapar K, Kovacs K, Scheithauer BW, et al. Proliferative activity and invasiveness among pituitary adenomas and carcinomas: An analysis using the MIB-1 antibody. Neurosurgery. 1996;38:99–106; discussion 106–107. 14. Thapar K, Scheithauer BW, Kovacs K, et al. p53 expression in pituitary adenomas and carcinomas: Correlation with invasiveness and tumor growth fractions. Neurosurgery. 1996;38:765– 770; discussion 770–771. 15. Kovacs K, Scheithauer BW, Horvath E, et al. The World Health Organization classification of adenohypophysial neoplasms. A proposed five-tier scheme. Cancer. 1996;78:502–510. 16. Yu R, Melmed S. Oncogene activation in pituitary tumors. Brain Pathol. 2001;11:328–341. 17. Ezzat S. The role of hormones, growth factors and their receptors in pituitary tumorigenesis. Brain Pathol. 2001;11:356–370. 18. Turner HE, Harris AL, Melmed S, et al. Angiogenesis in endocrine tumors. Endocr Rev. 2003;24:600–632. 19. Faglia G, Spada A. Genesis of pituitary adenomas: State of the art. J Neurooncol. 2001;54:95–110. 20. Musat M, Vax VV, Borboli N, et al. Cell cycle dysregulation in pituitary oncogenesis. Front Horm Res. 2004;32:34–62. 21. Suhardja AS, Kovacs KT, Rutka JT. Molecular pathogenesis of pituitary adenomas: A review. Acta Neurochir (Wien). 1999;141:729–736. 22. Saez C, Japon MA, Ramos-Morales F, et al. hpttg is overexpressed in pituitary adenomas and other primary epithelial neoplasias. Oncogene. 1999;18:5473–5476. 23. Zhang X, Horwitz GA, Heaney AP, et al. Pituitary tumor transforming gene (PTTG) expression in pituitary adenomas. J Clin Endocrinol Metab. 1999;84:761–767. 24. Alexander JM. Tumor suppressor loss in pituitary tumors. Brain Pathol. 2001;11:342–355. 25. Tzanela M. Dynamic tests and basal values for defining active acromegaly. Neuroendocrinology. 2006;83:200–204. 26. Simard M. The biochemical investigation of Cushing syndrome. Neurosurg Focus. 2004;16(4):E4. 27. Dinc H, Esen F, Demirci A, et al. Pituitary dimensions and volume measurements in pregnancy and post partum. MR assessment. Acta Radiol. 1998;39:64–69. 28. Castillo M. Pituitary gland: Development, normal appearances, and magnetic resonance imaging protocols. Top Magn Reson Imaging. 2005;16:259–268. 29. Amar AP, Weiss MH. Pituitary anatomy and physiology. Neurosurg Clin N Am. 2003;14:11–23, v. 30. Cottier JP, Destrieux C, Brunereau L, et al. Cavernous sinus invasion by pituitary adenoma: MR imaging. Radiology. 2000;215:463–469. 31. Knosp E, Steiner E, Kitz K, et al. Pituitary adenomas with invasion of the cavernous sinus space: A magnetic resonance imaging classification compared with surgical findings. Neurosurgery. 1993;33:610–617; discussion 617–618. 32. Saeki N, Yamaura A, Numata T, et al. Bone window CT evaluation of the nasal cavity for the transsphenoidal approach. Br J Neurosurg. 1999;13:285–289.
569
33. Jagannathan J, Prevedello DM, Ayer VS, et al. Computerassisted frameless stereotaxy in transsphenoidal surgery at a single institution: Review of 176 cases. Neurosurg Focus. 2006;20(2):E9. 34. Kucharczyk W, Bishop JE, Plewes DB, et al. Detection of pituitary microadenomas: Comparison of dynamic keyhole fast spin-echo, unenhanced, and conventional contrast-enhanced MR imaging. Am J Roentgenol. 1994;163:671–679. 35. Laws ER Jr. Pituitary surgery. Endocrinol Metab Clin North Am. 1987;16:647–665. 36. Bills DC, Meyer FB, Laws ER Jr, et al. A retrospective analysis of pituitary apoplexy. Neurosurgery. 1993;33:602–608; discussion 608–609. 37. Chuang CC, Chang CN, Wei KC, et al. Surgical treatment for severe visual compromised patients after pituitary apoplexy. J Neurooncol. 2006;80:39–47. 38. Ebersold MJ, Laws ER, Jr., Scheithauer BW, et al. Pituitary apoplexy treated by transsphenoidal surgery. A clinicopathological and immunocytochemical study. J Neurosurg. 1983;58:315–320. 39. Laws ER, Jr., Trautmann JC, Hollenhorst RW, Jr., Transsphenoidal decompression of the optic nerve and chiasm. Visual results in 62 patients. J Neurosurg. 1977;46:717–722. 40. Turner HE, Adams CB, Wass JA. Trans-sphenoidal surgery for microprolactinoma: An acceptable alternative to dopamine agonists? Eur J Endocrinol. 1999;140:43–47. 41. Amar AP, Couldwell WT, Chen JC, et al. Predictive value of serum prolactin levels measured immediately after transsphenoidal surgery. J Neurosurg. 2002;97:307–314. 42. Couldwell WT, Rovit RL, Weiss MH. Role of surgery in the treatment of microprolactinomas. Neurosurg Clin N Am. 2003;14:89– 92, vii. 43. Wang MY, Weiss MH. Is there a role for surgery for microprolactinomas? Semin Neurosurg. 2001;12:289–294. 44. Wolfsberger S, Czech T, Vierhapper H, et al. Microprolactinomas in males treated by transsphenoidal surgery. Acta Neurochir (Wien). 2003;145:935–940; discussion 940–941. 45. Liu JK, Couldwell WT. Contemporary management of prolactinomas. Neurosurg Focus. 2004;16(2):E2. 46. Nomikos P, Buchfelder M, Fahlbusch R. The outcome of surgery in 668 patients with acromegaly using current criteria of biochemical “cure.” Eur J Endocrinol. 2005;152:379– 387. 47. Ross DA, Wilson CB. Results of transsphenoidal microsurgery for growth hormone-secreting pituitary adenoma in a series of 214 patients. J Neurosurg. 1988;68:854–867. 48. Wass J. Debulking of pituitary adenomas improves hormonal control of acromegaly by somatostatin analogues. Eur J Endocrinol. 2005;152:693–694. 49. Grisoli F, Leclercq T, Jaquet P, et al. Transsphenoidal surgery for acromegaly—long-term results in 100 patients. Surg Neurol. 1985;23:513–519. 50. Laws ER Jr, Piepgras DG, Randall RV, et al. Neurosurgical management of acromegaly. Results in 82 patients treated between 1972 and 1977. J Neurosurg. 1979;50:454–461. 51. Sheaves R, Jenkins P, Blackburn P, et al. Outcome of transsphenoidal surgery for acromegaly using strict criteria for surgical cure. Clin Endocrinol (Oxf). 1996;45:407–413. 52. Colao A, Attanasio R, Pivonello R, et al. Partial surgical removal of growth hormone-secreting pituitary tumors enhances the response to somatostatin analogs in acromegaly. J Clin Endocrinol Metab. 2006;91:85–92. 53. Petrossians P, Borges-Martins L, Espinoza C, et al. Gross total resection or debulking of pituitary adenomas improves hormonal control of acromegaly by somatostatin analogs. Eur J Endocrinol. 2005;152:61–66. 54. Flitsch J, Ludecke DK, Knappe UJ, et al. Correlates of longterm hypocortisolism after transsphenoidal microsurgery for Cushing’s disease. Exp Clin Endocrinol Diabetes. 1999;107:183– 189. 55. Hofmann BM, Fahlbusch R. Treatment of Cushing’s disease: A retrospective clinical study of the latest 100 cases. Front Horm Res. 2006;34:158–184.
570
Hornyak and Couldwell
56. Knappe UJ, Ludecke DK. Persistent and recurrent hypercortisolism after transsphenoidal surgery for Cushing’s disease. Acta Neurochir Suppl. 1996;65:31–34. 57. Mampalam TJ, Tyrrell JB, Wilson CB. Transsphenoidal microsurgery for Cushing disease. A report of 216 cases. Ann Intern Med. 1988;109:487–493. 58. Sheehan JM, Lopes MB, Sheehan JP, et al. Results of transsphenoidal surgery for Cushing’s disease in patients with no histologically confirmed tumor. Neurosurgery. 2000;47:33–36; discussion 37–39. 59. Beckers A, Abs R, Mahler C, et al. Thyrotropin-secreting pituitary adenomas: Report of seven cases. J Clin Endocrinol Metab. 1991;72:477–483. 60. Snyder PJ. Extensive personal experience: Gonadotroph adenomas. J Clin Endocrinol Metab. 1995;80:1059–1061. 61. Ebersold MJ, Quast LM, Laws ER Jr, et al. Long-term results in transsphenoidal removal of nonfunctioning pituitary adenomas. J Neurosurg. 1986;64:713–719. 62. Fainstein Day P, Guitelman M, Artese R, et al. Retrospective multicentric study of pituitary incidentalomas. Pituitary. 2004;7:145–148. 63. Molitch ME. Approach to the incidentally discovered pituitary mass. Cancer Treat Res. 1997;89:73–90. 64. Molitch ME. Pituitary incidentalomas. Endocrinol Metab Clin North Am. 1997;26:725–740. 65. Nishizawa S, Ohta S, Yokoyama T, et al. Therapeutic strategy for incidentally found pituitary tumors (“pituitary incidentalomas”). Neurosurgery. 1998;43:1344–1348; discussion 1348–1350. 66. Oyama K, Sanno N, Tahara S, et al. Management of pituitary incidentalomas: According to a survey of pituitary incidentalomas in Japan. Semin Ultrasound CT MR. 2005;26:47–50. 67. Sanno N, Oyama K, Tahara S, et al. A survey of pituitary incidentaloma in Japan. Eur J Endocrinol. 2003;149:123–127. 68. Arita K, Tominaga A, Sugiyama K, et al. Natural course of incidentally found nonfunctioning pituitary adenoma, with special reference to pituitary apoplexy during follow-up examination. J Neurosurg. 2006;104:884–891. 69. Nielsen EH, Lindholm J, Bjerre P, et al. Frequent occurrence of pituitary apoplexy in patients with non-functioning pituitary adenoma. Clin Endocrinol (Oxf). 2006;64:319–322. 70. Vance ML. Perioperative management of patients undergoing pituitary surgery. Endocrinol Metab Clin North Am. 2003;32:355–365. 71. Inder WJ, Hunt PJ. Glucocorticoid replacement in pituitary surgery: Guidelines for perioperative assessment and management. J Clin Endocrinol Metab. 2002;87:2745–2750. 72. Gokalp HZ, Deda H, Attar A, et al. The neurosurgical management of prolactinomas. J Neurosurg Sci. 2000;44:128–132. 73. Brada M, Rajan B, Traish D, et al. The long-term efficacy of conservative surgery and radiotherapy in the control of pituitary adenomas. Clin Endocrinol (Oxf). 1993;38:571–578. 74. Kokubo M, Sasai K, Shibamoto Y, et al. Long-term results of radiation therapy for pituitary adenoma. J Neurooncol. 2000; 47:79–84. 75. McCollough WM, Marcus RB Jr, Rhoton Al Jr, et al. Long-term follow-up of radiotherapy for pituitary adenoma: The absence of late recurrence after greater than or equal to 4500 cGy. Int J Radiat Oncol Biol Phys. 1991;21:607–614. 76. McCord MW, Buatti JM, Fennell EM, et al. Radiotherapy for pituitary adenoma: Long-term outcome and sequelae. Int J Radiat Oncol Biol Phys. 1997;39:437–444. 77. Movsas B, Movsas TZ, Steinberg SM, et al. Long-term visual changes following pituitary irradiation. Int J Radiat Oncol Biol Phys. 1995;33:599–605. 78. Park P, Chandler WF, Barkan AL, et al. The role of radiation therapy after surgical resection of nonfunctional pituitary macroadenomas. Neurosurgery. 2004;55:100–106; discussion 106–107. 79. Rush S, Cooper PR. Symptom resolution, tumor control, and side effects following postoperative radiotherapy for pituitary macroadenomas. Int J Radiat Oncol Biol Phys. 1997;37:1031– 1034.
80. Tsang RW, Brierley JD, Panzarella T, et al. Role of radiation therapy in clinical hormonally-active pituitary adenomas. Radiother Oncol. 1996;41:45–53. 81. Zierhut D, Flentje M, Adolph J, et al. External radiotherapy of pituitary adenomas. Int J Radiat Oncol Biol Phys. 1995;33:307– 314. 82. Hughes MN, Llamas KJ, Yelland ME, et al. Pituitary adenomas: Long-term results for radiotherapy alone and post-operative radiotherapy. Int J Radiat Oncol Biol Phys. 1993;27:1035–1043. 83. Grigsby PW, Simpson JR, Stokes S, et al. Results of surgery and irradiation or irradiation alone for pituitary adenomas. J Neurooncol. 1988;6:129–134. 84. Estrada J, Boronat M, Mielgo M, et al. The long-term outcome of pituitary irradiation after unsuccessful transsphenoidal surgery in Cushing’s disease. N Engl J Med. 1997;336:172–177. 85. Fisher BJ, Gaspar LE, Noone B. Radiation therapy of pituitary adenoma: Delayed sequelae. Radiology. 1993;187:843–846. 86. Barnett GH, Linskey ME, Adler JR, et al. Stereotactic radiosurgery—an organized neurosurgery-sanctioned definition. J Neurosurg. 2006;106:1–5. 87. Chen JC, Giannotta SL, Yu C, et al. Radiosurgical management of benign cavernous sinus tumors: Dose profiles and acute complications. Neurosurgery. 2001;48:1022–1030; discussion 1030– 1032. 88. Couldwell WT, Kan P, Liu JK, et al. Decompression of cavernous sinus meningioma for preservation and improvement of cranial nerve function. Technical note. J Neurosurg. 2006;105:148– 152. 89. Couldwell WT, Rosenow JM, Rovit RL, et al. Hypophysopexy technique for radiosurgical treatment of cavernous sinus pituitary adenoma. Pituitary. 2002;5:169–173. 90. Liu JK, Schmidt MH, MacDonald JD, et al. Hypophysial transposition (hypophysopexy) for radiosurgical treatment of pituitary tumors involving the cavernous sinus. Technical note. Neurosurg Focus. 2003;14(5):e11. 91. Laws ER, Sheehan JP, Sheehan JM, et al. Stereotactic radiosurgery for pituitary adenomas: A review of the literature. J Neurooncol. 2004;69:257–272. 92. Sheehan JP, Niranjan A, Sheehan JM, et al. Stereotactic radiosurgery for pituitary adenomas: An intermediate review of its safety, efficacy, and role in the neurosurgical treatment armamentarium. J Neurosurg. 2005;102:678–691. 93. Laws ER Jr, Vance ML. Conventional radiotherapy for pituitary tumors. Neurosurg Clin N Am. 2000;11:617–625. 94. Picozzi P, Losa M, Mortini P, et al. Radiosurgery and the prevention of regrowth of incompletely removed nonfunctioning pituitary adenomas. J Neurosurg. 2005;102 (Suppl):71–74. 95. Colin P, Jovenin N, Delemer B, et al. Treatment of pituitary adenomas by fractionated stereotactic radiotherapy: A prospective study of 110 patients. Int J Radiat Oncol Biol Phys. 2005;62:333– 341. 96. Baumert BG, Norton IA, Lomax AJ, et al. Dose conformation of intensity-modulated stereotactic photon beams, proton beams, and intensity-modulated proton beams for intracranial lesions. Int J Radiat Oncol Biol Phys. 2004;60:1314–1324. 97. Milker-Zabel S, Zabel A, Huber P, et al. Stereotactic conformal radiotherapy in patients with growth hormone-secreting pituitary adenoma. Int J Radiat Oncol Biol Phys. 2004;59:1088–1096. 98. Minniti G, Traish D, Ashley S, et al. Fractionated stereotactic conformal radiotherapy for secreting and nonsecreting pituitary adenomas. Clin Endocrinol (Oxf). 2006;64:542–548. 99. Paek SH, Downes MB, Bednarz G, et al. Integration of surgery with fractionated stereotactic radiotherapy for treatment of nonfunctioning pituitary macroadenomas. Int J Radiat Oncol Biol Phys. 2005;61:795–808. 100. Besser M. Criteria for medical as opposed to surgical treatment of prolactinomas. Acta Endocrinol (Copenh). 1993;129 (Suppl 1):27–30. 101. Bevan JS, Webster J, Burke CW, et al. Dopamine agonists and pituitary tumor shrinkage. Endocr Rev. 1992;13:220–240. 102. Ciccarelli E, Camanni F. Diagnosis and drug therapy of prolactinoma. Drugs. 1996;51:954–965.
Chapter 39: Pituitary Adenomas 103. Cunnah D, Besser M. Management of prolactinomas. Clin Endocrinol (Oxf). 1991;34:231–235. 104. Weiss MH, Wycoff RR, Yadley R, et al. Bromocriptine treatment of prolactin-secreting tumors: Surgical implications. Neurosurgery. 1983;12:640–642. 105. Weiss MH. Medical and surgical management of functional pituitary tumors. Clin Neurosurg. 1981;28:374–383. 106. Molitch ME. Pharmacologic resistance in prolactinoma patients. Pituitary. 2005;8:43–52. 107. Morange I, Barlier A, Pellegrini I, et al. Prolactinomas resistant to bromocriptine: Long-term efficacy of quinagolide and outcome of pregnancy. Eur J Endocrinol. 1996;135:413–420. 108. Colao A, Di Sarno A, Cappabianca P, et al. Withdrawal of longterm cabergoline therapy for tumoral and nontumoral hyperprolactinemia. N Engl J Med. 2003;349:2023–2033. 109. Couldwell WT, Weiss MH, Laws ER, Jr., Prolactinomas. N Engl J Med. 2004;350:1054–1057; author reply 1054–1057. 110. Tyrrell JB, Lamborn KR, Hannegan LT, et al. Transsphenoidal microsurgical therapy of prolactinomas: Initial outcomes and long-term results. Neurosurgery. 1999;44:254–261; discussion 261–263. 111. Thorner MO. Prolactinoma. West J Med. 1983;139:703–705. 112. Vance ML, Thorner MO. Prolactinomas. Endocrinol Metab Clin North Am. 1987;16:731–753. 113. Blakey DC, Westwood FR, Walker M, et al. Antitumor activity of the novel vascular targeting agent ZD6126 in a panel of tumor models. Clin Cancer Res. 2002;8:1974–1983. 114. McIntyre DJ, Robinson SP, Howe FA, et al. Single dose of the antivascular agent, ZD6126 (N-acetylcolchinol-O-phosphate), reduces perfusion for at least 96 hours in the GH3 prolactinoma rat tumor model. Neoplasia. 2004;6:150–157. 115. Swearingen B, Barker FG II, Katznelson L, et al. Long-term mortality after transsphenoidal surgery and adjunctive therapy for acromegaly. J Clin Endocrinol Metab. 1998;83:3419–3426. 116. Ezzat S, Snyder PJ, Young WF, et al. Octreotide treatment of acromegaly. A randomized, multicenter study. Ann Intern Med. 1992;117:711–718. 117. van der Lely AJ, Hutson RK, Trainer PJ, et al. Long-term treatment of acromegaly with pegvisomant, a growth hormone receptor antagonist. Lancet. 2001;358:1754–1759. 118. Colao A, Ferone D, Marzullo P, et al. Effect of different dopaminergic agents in the treatment of acromegaly. J Clin Endocrinol Metab. 1997;82:518–523. 119. Cozzi R, Attanasio R, Barausse M, et al. Cabergoline in acromegaly: A renewed role for dopamine agonist treatment? Eur J Endocrinol. 1998;139:516–521. 120. Freda PU, Reyes CM, Nuruzzaman AT, et al. Cabergoline therapy of growth hormone & growth hormone/prolactin secreting pituitary tumors. Pituitary. 2004;7:21–30. 121. Dimaraki EV, Symons KV, Barkan AL. Raloxifene decreases serum IGF-I in male patients with active acromegaly. Eur J Endocrinol. 2004;150:481–487. 122. van der Hoek J, Hofland LJ, Lamberts SW. Novel subtype specific and universal somatostatin analogues: Clinical potential and pitfalls. Curr Pharm Des. 2005;11:1573–1592. 123. Jaquet P, Gunz G, Saveanu A, et al. BIM-23A760, a chimeric molecule directed towards somatostatin and dopamine receptors, vs universal somatostatin receptors ligands in GHsecreting pituitary adenomas partial responders to octreotide. J Endocrinol Invest. 2005;28:21–27. 124. Jaquet P, Saveneau A, Barlier A. New SRIF analogs in the control of human pituitary adenomas: Perspectives. J Endocrinol Invest 2005;28:14–18. 125. Vance ML. Medical treatment of functional pituitary tumors. Neurosurg Clin N Am. 2003;14:81–87. 126. Oyesiku NM. Multimodality treatment of pituitary adenomas. Clin Neurosurg 2005;52:234–242. 127. Nagesser SK, van Seters AP, Kievit J, et al. Long-term results of total adrenalectomy for Cushing’s disease. World J Surg. 2000;24:108–113. 128. Heaney AP, Fernando M, Yong WH, et al. Functional PPAR-gamma receptor is a novel therapeutic target for
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
140.
141.
142. 143.
144.
145. 146. 147.
148.
149. 150.
151.
571
ACTH-secreting pituitary adenomas. Nat Med. 2002;8:1281– 1287. Heaney AP, Fernando M, Melmed S. PPAR-gamma receptor ligands: Novel therapy for pituitary adenomas. J Clin Invest. 2003;111:1381–1388. Ambrosi B, Dall’Asta C, Cannavo S, et al. Effects of chronic administration of PPAR-gamma ligand rosiglitazone in Cushing’s disease. Eur J Endocrinol. 2004;151:173–178. Giraldi FP, Scaroni C, Arvat E, et al. Effect of protracted treatment with rosiglitazone, a PPARgamma agonist, in patients with Cushing’s disease. Clin Endocrinol (Oxf). 2006;64:219– 224. Hull SS, Sheridan B, Atkinson AB. Pre-operative medical therapy with rosiglitazone in two patients with newly diagnosed pituitary-dependent Cushing’s syndrome. Clin Endocrinol (Oxf). 2005;62:259–261. Bogazzi F, Ultimieri F, Raggi F, et al. PPARgamma inhibits GH synthesis and secretion and increases apoptosis of pituitary GHsecreting adenomas. Eur J Endocrinol. 2004;150:863–875. Gruszka A, Kunert-Radek J, Pawlikowski M. Rosiglitazone, PPAR-gamma receptor ligand, decreases the viability of rat prolactin-secreting pituitary tumor cells in vitro. Neuro Endocrinol Lett. 2005;26:51–54. Mannavola D, Persani L, Vannucchi G, et al. Different responses to chronic somatostatin analogues in patients with central hyperthyroidism. Clin Endocrinol (Oxf). 2005;62:176–181. Kaltsas GA, Mukherjee JJ, Plowman PN, et al. The role of cytotoxic chemotherapy in the management of aggressive and malignant pituitary tumors. J Clin Endocrinol Metab. 1998;83:4233–4238. Laws ER, Jr., Morris AM, Maartens N. Gliadel for pituitary adenomas and craniopharyngiomas. Neurosurgery. 2003;53:255– 269; discussion 259–260. Lim S, Shahinian H, Maya MM, et al. Temozolomide: A novel treatment for pituitary carcinoma. Lancet Oncol. 2006;7:518– 520. Fadul CE, Kominsky AL, Meyer LP, et al. Long-term response of pituitary carcinoma to temozolomide. Report of two cases. J Neurosurg. 2006;105:621–626. Kovacs K, Horvath E, Syro LV, et al. Temozolomide therapy in a man with an aggressive prolactin-secreting pituitary neoplasm: Morphological findings. Hum Pathol. 2006;38:185–189. Ciric I, Ragin A, Baumgartner C, et al. Complications of transsphenoidal surgery: Results of a national survey, review of the literature, and personal experience. Neurosurgery. 1997;40:225–236; discussion 236–237. Giovanelli M, Losa M, Mortini P, et al. Surgical results in microadenomas. Acta Neurochir Suppl. 1996;65:11–12. Rees DA, Hanna FW, Davies JS, et al. Long-term follow-up results of transsphenoidal surgery for Cushing’s disease in a single centre using strict criteria for remission. Clin Endocrinol (Oxf). 2002;56:541–551. Semple PL, Laws ER Jr Complications in a contemporary series of patients who underwent transsphenoidal surgery for Cushing’s disease. J Neurosurg. 1999;91:175–179. Laws ER Jr, Kern EB. Complications of trans-sphenoidal surgery. Clin Neurosurg. 1976;23:401–416. Wilson CB, Dempsey LC. Transsphenoidal microsurgical removal of 250 pituitary adenomas. J Neurosurg. 1978;48:13–22. Couldwell WT, Kan P, Weiss MH. Simple closure following transsphenoidal surgery. Technical note. Neurosurg Focus. 2006;20(3):E11. Nemergut EC, Zuo Z, Jane JA, Jr., et al. Predictors of diabetes insipidus after transsphenoidal surgery: A review of 881 patients. J Neurosurg. 2005;103:448–454. Barrow DL, Tindall GT. Loss of vision after transsphenoidal surgery. Neurosurgery. 1990;27:60–68. van Aken MO, de Marie S, van der Lely AJ, et al. Risk factors for meningitis after transsphenoidal surgery. Clin Infect Dis. 1997;25:852–856. Laws ER, Jr. Vascular complications of transsphenoidal surgery. Pituitary. 1999;2:163–170.
572
Hornyak and Couldwell
152. Tan LK, Jones RA. Nasal complications of the direct transnasal approach to the pituitary fossa. Br J Neurosurg. 1995;9:739– 742. 153. Mortini P, Losa M, Barzaghi R, et al. Results of transsphenoidal surgery in a large series of patients with pituitary adenoma. Neurosurgery. 2005;56:1222–1233; discussion 1233. 154. Dekkers OM, Pereira AM, Roelfsema F, et al. Observation alone after transsphenoidal surgery for nonfunctioning pituitary macroadenoma. J Clin Endocrinol Metab. 2006;91:1796– 1801. 155. Chen JC, Amar AP, Choi S, et al. Transsphenoidal microsurgical treatment of Cushing disease: Postoperative assessment of surgical efficacy by application of an overnight low-dose dexamethasone suppression test. J Neurosurg. 2003;98:967–973. 156. Esposito V, Santoro A, Minniti G, et al. Transsphenoidal adenomectomy for GH-, PRL- and ACTH- secreting pituitary tumours: Outcome analysis in a series of 125 patients. Neurol Sci 2004;25:251–256.
157. Guilhaume B, Bertagna X, Thomsen M, et al. Transsphenoidal pituitary surgery for the treatment of Cushing’s disease: Results in 64 patients and long term follow-up studies. J Clin Endocrinol Metab. 1988;66:1056–1064. 158. Laws ER, Vance ML, Thapar K. Pituitary surgery for the management of acromegaly. Horm Res. 2000;53 (suppl 3):71–75. 159. Shimon I, Ram Z, Cohen ZR, et al. Transsphenoidal surgery for Cushing’s disease: Endocrinological follow-up monitoring of 82 patients. Neurosurgery. 2002;51:57–61; discussion 61–62. 160. Hashimoto N, Handa H, Yamashita J, et al. Long-term followup of large or invasive pituitary adenomas. Surg Neurol. 1986;25:49–54. 161. Abe T, Ludecke DK. Recent results of secondary transnasal surgery for residual or recurring acromegaly. Neurosurgery. 1998;42:1013–1021; discussion 1021–1022. 162. Trepp R, Stettler C, Zwahlen M, et al. Treatment outcomes and mortality of 94 patients with acromegaly. Acta Neurochir (Wien). 2005;147:243–251; discussion 250–251.
40 Craniopharyngioma: Neurosurgical Management Douglas James Cook and James T. Rutka
pose tissue have been proposed as an underlying mechanism (12). 4. Increased intracranial pressure: Headache, nausea and vomiting, and papilloedema occur in 15% to 30% of patients when a craniopharyngioma blocks CSF flow with resulting obstructive hydrocephalus. This syndrome of raised intracranial pressure (ICP) is common with large tumors or tumors with large cystic components that extend upwards to fill the third ventricle (13). 5. Disturbance in behavior or mentation: Altered cognition or behavior is usually due to mass effect on the frontotemporal lobes secondary to a large tumor or hydrocephalus. Psychiatric symptoms may occur in 40% of adults and less than 10% of children, who may be more difficult to assess in this regard. Altered memory and mentation, and depression predict a poor prognosis (4,5,11). Apathy, incontinence, hypersomnia, and even Korsakoff syndrome may also be noted. Preoperative neuro-psychological testing is recommended to document these disturbances.
INCIDENCE AND EPIDEMIOLOGY Craniopharyngioma is a relatively rare lesion comprising an estimated 3% of intracranial tumors (1,2). There is a bimodal distribution of craniopharyngioma by age peaking in the 5to 14-year-old and 50- to 74-year-old groups. The diagnosis is more common in the pediatric group, comprising 10% of pediatric brain tumors diagnosed annually in the United States. The incidence of craniopharyngioma in North America is 0.13 per 100,000 (1). The Childhood Cancer Registry of Piedmont, Italy, estimated an incidence of 0.14 per 100,000 (2). A higher incidence has been observed in Asian countries, reported as high as 0.53 per 100,000 in one Japanese study (2,3). Craniopharyngioma can present as one of several clinical syndromes based on its origin and growth pattern. Data from Carmel (4) and Banna (5,6) show that children and adults have dissimilar clinical syndromes: Children present with a higher proportion of endocrine dysfunction, progressive visual loss, symptoms of raised intracranial pressure, and papilloedema, whereas adults more uniformly have visual impairment. Clinical presentation can be subdivided into five categories.
Intraventricular rupture of a cystic craniopharyngioma has been reported, and can result in acute clinical deterioration (14,15). This rare phenomenon can produce headache, fever, nuchal rigidity, and obtundation secondary to chemical meningitis.
1. Endocrine dysfunction: Short stature and delay in onset of puberty can be seen in children, as well as diabetes insipidus and hypothyroidism in children and adults. Panhypopituitarism or selective pituitary hormone deficiency due to compression or invasion of the pituitary gland and infundibulum can be found on baseline assessments of plasma ACTH, serum and urine cortisol, electrolytes and osmolarity, plasma TSH, T4 FSH, LH, GH, testosterone, estradiol, and prolactin. With purely intrasellar tumors, older patients may present with amenorrhea/galactorrhea and infertility due to increased levels of serum prolactin. Formal neuroendocrine testing consisting of evaluation of pituitary or target-organ hormones and response of the hypothalamic–pituitary axis to intravenously administered stimulators should be undertaken preoperatively and followed postoperatively (7,8). 2. Visual disturbance: Visual complaints such as decreased visual acuity, visual field deficits (commonly bitemporal hemianopia), and optic atrophy are present in 37% to 70% of children and 85% of adults at presentation (4–6,9). Often reversible after surgery, visual loss or field cuts may be due to compression-related disruption of axoplasmic flow and demyelination in the anterior visual pathways (10). Formal neuro-ophthalmologic testing should be obtained as a preoperative baseline of visual function in all patients. 3. Hypothalamic dysfunction: Hyperphagia and obesity can occur, but is often most pronounced postoperatively (11). Elevated serum leptin levels and disturbed feedback mechanisms from hypothalamic leptin receptors to adi-
PATHOLOGY In 1838 Rathke described the embryology of the pituitary gland and identified the ectodermal stomadeal diverticulum as the major contributor to the anterior pituitary stalk (16). In 1899, Mott and Baratt described a third ventricular tumor that was related to the hypophyseal–pharyngeal duct (17), a squamous epithelialized structure between pituitary gland and pharynx that Luschka had originally described in 1860 (18). Erdheim described the neuroembryological origin of craniopharyngioma in 1904 when he attributed the formation of this tumor type to incomplete involution of the hypophyseal–pharyngeal duct. Cushing coined the term “craniopharyngioma,” to unify the prior terminology of this pathology which included hypophyseal–pharyngeal duct tumor, Rathke pouch tumor, adamantimoma, and ameloblastoma (19). The embryological formation of the hypophyseal– pharyngeal duct occurs at the fourth week of gestation when the neurohypophysis and buccopharyngeal membrane come into contact (20,21). The neurohypophysis arises from an outpouching at the junction of the diencephalon and mesencephalon which forms a tubular structure as the neural tube flexes. The buccopharyngeal membrane forms an ectodermal outpouching, Rathke pouch, that projects to the neurohypophysis. These two structures fuse at gestational age 8 573
574
Cook and Rutka
weeks and form a tubular structure that goes on to collapse into the solid anterior infundibulum (21). There are two distinct histologic pathologies in craniopharynigoma, adamantinomatous, and papillary. The underlying mechanism behind each pathology is unique, but both are based on incomplete closure of the hypophyseal– pharyngeal duct. Adamantinomatous craniopharyngioma, named for its similarities to dental tissue, is the most common form of craniopharyngioma in children. Adamantinomatous craniopharyngioma is thought to arise from persistent elements of the buccopharyngeal membrane or hypophyseal– stomadeal duct that represent tissue destined to be tooth primordia (22,23). This odontogenic tissue undergoes metaplasia resulting in tumor formation. The papillary craniopharyngioma is most common in adults and thought to arise from metaplasia of the squamous epithelial “rest," remnants of the stomadeum destined to become the buccal mucosa (24–27). On gross examination, adamantinomatous craniopharyngioma is cystic with brown, turbid, cholesterol-rich fluid. The solid components of the tumor may contain calcified components and disorganized squamous component. Microscopic examination includes pallisading columnar cells arranged in irregular shapes and/or sheets and whorls around a loose stellate epithelium (28,29). Calcification and bony formation are common among eosinophilic islands of desquamating epithelium. A gliotic capsule at the junction of tumor and brain is common with Rosenthal fibers (28). Fibrohistiocytic reaction and cholesterol clefts are frequently seen. In contrast, papillary craniopharyngioma is typically a solid tumor with a low propensity for cyst formation, calcification, and lacking cholesterol deposits and the turbid cyst fluid. Microscopic examination reveals a differentiated squamous epithelium arranged in pseudopapillae or papillae around a fibrovascular core (29). Epithelial immunohistochemical markers such as cytokeratin-7 and epithelial membrane antigen are strongly positive (30). There is no reliable molecular marker for either type of craniopharyngioma; however, several studies have identified chromosomal losses and gains and mutated oncogenes in these lesions that may be significant on a case by case basis (31–34). Rickert and Paulus published a series of 29 carniopharyngiomas (20 admantinomatous and 9 papillary) where there were no chromosomal abnormalities by comparative genomic hybridization (35). Yoshimoto et al. concluded that the chromosomal losses and gains identified by genomic hybridization in pediatric adamantinomatous craniopharyngiomas do not contribute to tumorigenesis (36). The presentation of patients with craniopharyngioma is highly variable based on its growth pattern. Because of the clinical and surgical importance of the different growth patterns, Hoffman proposed the following classification: (i) Sellar craniopharyngioma confined to the sella or protruding only a short distance above the diaphragma. (ii) Prechiasmatic craniopharyngioma growing forward and pushing the optic chiasm and anterior cerebral arteries superior and posterior. (iii) Retrochiasmatic craniopharyngioma displacing the optic chiasm anterior and growing back into the third ventricle. (iv) Giant craniopharyngioma with varying growth patterns (37,38). Wang et al. have proposed a revised classification of craniopharyngioma that emphasizes the relationship of the tumor to the diaphragma (39). The first division is subdiaphragmatic versus supradiaphragmatic. Subdiaphragmatic lesions are then divided into subdiaphragmatic with competent diaphragma sellae versus subdiaphragmatic with incompetent diaphragma sellae. Subdiaphragmatic lesions with a
competent diaphragma will grow to distend the diaphragm and dura maintaining a “capsule” over the superior elements of the lesion. This lesion configuration prevents adhesions between tumor and adjacent neural structures. These lesions push upwards on the optic chiasm and anterior cerebral vessels. Subdiaphragmatic lesions with an incompetent diaphragma will acquire a “dumbbell” configuration as the growth will spill through the aperture of the diaphragma into the suprasellar space. The diaphragm forms the waist of the lesion. These lesions are oriented in the retrochiasmatic direction and continued growth results in anterior displacement of the optic chiasm, adhesions to superior neural elements, particularly the hypothalamus, and growth into the third ventricle through its thin floor. Supradiaphragmatic lesions have no intrasellar component and grow in the retrochiasmatic direction as well as laterally into adjacent CSF cisternae. These lesions can adhere to adjacent neural structures including hypothalamus, optic nerve and tract, floor of the third ventricle, and lower cranial nerves. These lesions have a propensity to erode through the thin floor of the third ventricle causing hydrocephalus.
STAGING Staging of craniopharyngioma includes a thorough biochemical workup, diagnostic imaging, and neuropsychological testing. Biochemical workup consists of serum levels of GH, LH/FSH, ACTH, TSH, and Prolactin. Diabetes insipidus must also be ruled out by taking a history of urinary output, thirst, and measuring serum sodium, osmolality and urine-specific gravity. Up to 85% of patients will have one or more abnormalities in this panel (40). Specifically, GH deficiency occurs in 35% to 95% of craniopharyngioma patients, LH/FSH deficiency in 38% to 82%, ACTH deficiency in 21% to 62%, TSH deficiency in 21% to 42%, and diabetes insipidus in 6% to 38% (41). A thorough imaging workup includes CT scanning with and without contrast and reconstruction of the sella and skull base, MRI with and without gadolinium and MR angiography. Skull X-rays are no longer standard imaging, but can reveal enlargement and/or destruction of the sella with intra- and suprasellar calcification associated with the tumor. Angiography was also utilized historically for localization of the tumor and involvement of vascular structures, in particular the anterior cerebral vessel A1 and A2 segments; however, this has been largely replaced by noninvasive vascular imaging modalities. CT imaging reveals the hallmark calcification present in 50% of adult craniopharyngiomas and the vast majority of pediatric cases (Fig. 1) (42–44). CT effectively demonstrates hydrocephalus in patients presenting with acute deterioration. Reconstructions of the skull base and sella demonstrate the bony involvement of the tumor including expansion of the sella, erosion of dorsum sellae and clinoids, and projection into the sinuses where applicable. Finally, CT angiography demonstrates the proximity and involvement of the tumor with vital vascular structures, particularly in the carotid and anterior circulation. For a detailed understanding of the relationship of the tumor with neural structures, MRI is required (Fig. 2) (42,44,45). MRI T1 sequence images in the axial, sagittal, and coronal planes delineate tumor margins with relationships to pituitary, pituitary stalk, optic chiasm and tracts, hypothalamus, and third ventricle. Gadolinium administration reveals
Chapter 40: Craniopharyngioma: Neurosurgical Management Table 1
575
Typical Features of Craniopharyngiomas on Imaging Studies
1. Suprasellar mass, with solid nodule and associated cyst. A cyst is seen in the majority of craniopharyngiomas, with rates reported as high as 95%. Of these, the majority has both solid and cystic components, with only a minority being cystic alone (55). 2. Cyst contents typically higher attenuation than CSF 3. Nodular or rim calcification 4. Nodule and rim typically enhance with contrast agents 5. Heterogeneous mass on MRI. The cyst is typically hypointense on T1 and hyperintense on T2. Cysts that are hyperintense on T1 are likely due to blood degradation products, high protein concentration or both.
Figure 1 Axial CT of a 11-year-old male with history of right visual loss and headaches. CT shows densely calcified suprasellar lesion.
enhancement of nodular components of tumor and in cystic walls. T2 images in axial and sagittal planes demonstrate cystic components, typically hyperintense, and can delineate cyst and cyst wall from ventricle. MR angiography is a useful tool to assess arterial anatomy, displacement by, and association with tumor. Table 1 describes typical imaging findings for craniopharyngioma. Table 2 includes a differential diagnosis for lesions of the sella on imaging.
Figure 2 10-year-old boy with visual failure, headaches, and short stature. (Top left) Preoperative sagittal MRI showing large solid:cystic lesions invaginating into the third ventricle. (Top right) Postoperative sagittal MRI showing complete resection of the lesion. (Bottom left) Preoperative coronal MRI showing extensive nature of tumor before surgery. (Bottom right) Postoperative MRI showing complete resection of the lesion.
576
Cook and Rutka Table 2 Differential Diagnosis of Sellar Lesion on Imaging 1. Rathke cleft cyst 2. Dermoid cyst 3. Epidermoid cyst 4. Pituitary adenoma 5. Germinoma 6. Hamartoma 7. Suprasellar aneurysm 8. Arachnoid cyst 9. Suprasellar abcess 10. Langerhans cell histiocytosis 11. Sarcoidosis 12. Tuberculosis 13. Hypothalamic or optic glioma 14. Meningioma
TREATMENT The treatment of craniopharyngiomas is multifaceted, given their location, diverse clinical presentations, and the potential complications of treatment. Above all, the treatment goals are to preserve to the best extent possible hypothalamic, endocrine, and visual function. Treatment strategies need therefore to consider several distinct areas: Management of hydrocephalus. preoperative medical management, tumor resection. postoperative care, and adjuvant treatment.
Management of Hydrocephalus Hydrocephalus is a common presentation for patients with craniopharyngiomas. In a review of craniopharyngiomas at the Hospital for Sick Children in Toronto, 48% of patients presented with preoperative hydrocephalus (46); however, percentages as low as 16.7% in children and 13.3% in adults have also been reported (47). The level of obstruction is often the foramen of Monro, and therefore it is commonly seen with retrochiasmatic tumors (37). A prechiasmatic tumor may uncommonly lead to hydrocephalus, if it is large enough to do so on the basis of size alone. The most common method of treatment of hydrocephalus associated with craniopharyngiomas is operative treatment of the tumor itself. Treatment of hydrocephalus alone may be indicated if the patient presents in extremis due to increased intracranial pressure. One must consider, however, some of the complications associated with longterm CSF shunting in patients with craniopharyngioma. Tomita and McLone report two patients in whom preoperative shunting resulted in acute visual deterioration, including blindness in one patient, presumably due to upward herniation of the tumor after ventricular decompression (48). Patients that are shunted before or after surgical resection of the tumor may have a worse prognosis, possibly due to tumor size or extensive hypothalamic damage (49).
Preoperative Medical Management Perhaps more than in other patients with intracranial tumors, the medical management of the craniopharyngioma patient has great bearing on the overall outcome. In addition to undergoing extensive endocrinological investigations, specific attention must be paid to the preoperative, intraoperative, and postoperative endocrinological status of the patient. The use of steroids is necessary for patients presenting with adrenal dysfunction, and is also indicated for treatment of vasogenic edema associated with the tumor. Endocrinological assessments of craniopharyngioma patients suggest that
all patients should be presumed as deficient in ACTH and supplemented. Additional steroid coverage is not required preoperatively if patients are receiving dexamethasone for tumor control (50). Diabetes insipidus (DI) is a presenting symptom in 6% to 38% of patients (41) and has been reported to be present in 60% to 70% of postoperative craniopharyngiomas (51). In patients who present with DI, or in the rare instances of intraoperative development of DI, intraoperative continuous infusion of aqueous vasopressin may be necessary. More commonly however, DI is seen postoperatively in the second or third day, emphasizing the crucial role for postoperative water and electrolyte management in these patients (38). Sodium and electrolyte balance is directly tied to water balance, and these parameters need to be closely monitored. Intraoperative use of mannitol should be minimized in order to optimize fluid management in the patient. Intraoperative drainage of CSF through a ventriculostomy may be helpful in averting the need for mannitol, and will facilitate gentle brain retraction. It is not uncommon for seizures to occur in the postoperative period. These tend to be self-limited, and long-term seizure disorders are not commonly seen. We advocate the use of anticonvulsant medication, which can be tapered and discontinued if there is no seizure activity after 10 days (38).
Neurosurgical Approaches for Craniopharyngioma The ultimate goal in the treatment of craniopharyngiomas is to obliterate the tumor while avoiding endocrinological and visual morbidity. If this is not possible, the goal should be decompression of the optic chiasm and the ventricular system. The potential morbidity resulting from poorly executed aggressive surgery, including hyperphagia and morbid obesity, blindness or panhypopituitarism will have severe, life-long implications for the quality of life of the patient. It is not unexpected therefore that there is significant debate as to which approach is best suited for patients with craniopharyngioma. The main neurosurgical approaches are aimed at gross total resection, subtotal surgical resection combined with radiotherapy, and intracavitary therapy. The current approaches will be discussed, followed by the main arguments for the controversy. The neurosurgical approach to a craniopharyngioma is in large part dictated by the growth characteristics of the tumor (13,49,52). The most practical manner to classify craniopharyngiomas is with respect to their anatomical relationship to the chiasm. Therefore, the position of the tumor and its intimate relationship with the visual pathways in large part dictates the operative approach. Patient positioning is important as with any microsurgical operation, since detailed anatomical relationships are to be visualized. We have preferred a subfrontal approach in the majority of children with craniopharyngioma. We recognize, however, that several other neurosurgical approaches can be undertaken such as the unilateral pterional, subtemporal, and transpetrous approaches. For the subfrontal approach, the head is placed in pin fixation, unless the patient is younger than 3 years. Typically, the patient is placed supine with the head slightly extended and the nose pointing upwards. The head extension allows the frontal lobes to fall away from the anterior cranial fossa, minimizing the risk of retraction injury. We utilize a unilateral orbitotomy with an angled osteotomy across the midline (Fig. 3) to facilitate exposure of the anterior cranial base, and to minimize brain retraction. A unilateral dural opening is performed which is typically opened across the midline. The superior sagittal sinus is ligated far
Chapter 40: Craniopharyngioma: Neurosurgical Management
577
Figure 3 Artist’s depiction of operative approach for craniopharyngioma. Unilateral subfrontal approach. (Left) The right bone flap is carried just over the midline to the left. A unilateral orbital osteotomy is performed as depicted. The inferior and medial cut of the osteotomy is angled across the midline so as to have as much exposure at the cranial base in the midline as possible. (Right) With the bone flap removed, the dural opening can be performed as indicated by the hatched line. The dura is reflected inferiorly as far as possible to reveal the anterior cranial base. The operative procedure should be conducted with as little subfrontal brain retraction as possible.
anteriorly, approximately 4 cm above the orbital osteotomy to facilitate brain relaxation further. Once the dura is open, the operation shifts to the use of microsurgical techniques. Tumor removal is benefited by the use of the Cavitron. Use of frameless stereotaxy and intraoperative neuronavigation assist in localization and identification of landmarks in skull base procedures, but we have not found them to be uniformly useful.
Prechiasmatic and sellar lesions can be approached through a unilateral subfrontal approach, typically from the right (nondominant) side, or an anterior interhemispheric approach. These allow adequate exposure of the chiasmatic cistern, the optic apparatus, internal carotid arteries, and the tumor itself (Fig. 4) (38). Retrochiasmatic tumors are best approached through a subfrontal or combined subfrontal-pterional approach. For
Figure 4 (Left) Intraoperative exposure of retrochiasmatic craniopharyngioma showing left and right optic nerves leading to the chiasm. The left and right olfactory nerves are seen. Tumor is found beneath the chiasm and in the prechiasmatic space. Calcified tumor is also appreciated lateral to the left optic nerve as shown. (Right) At the end of the procedure, curved micromirrors can be used to inspect the blind corridors such as the subchiasmatic region (as shown), and beneath the optic tracts. One can frequently see the basilar apex with the posterior cerebral arteries behind the membrane of Lillequist (as shown), as well as the A1 and A2 segments of the anterior cerebral arteries. All these critical structures need to be identified and preserved throughout the procedure.
578
Cook and Rutka
Figure 5 A 14-year-old female with bitemporal hemianopsia, and weight gain. The lesion was removed via a trans-sphenoidal approach. (Left) Coronal MRI with contrast showing large sellar mass with significant suprasellar extension. (Right) Sagittal MRI with contrast showing expanded sella, with predominantly cystic contents.
retrochiasmatic craniopharyngiomas, the tumor needs to be delivered through the lamina terminalis if the subfrontal approach is utilized. The latter is achieved by identifying the A1 segments bilaterally, and gently elevating them above the optic chiasm (Fig. 4). Sellar lesions typically arise below the level of the diaphragma sellae and often lead to sellar enlargement (Fig. 5). The dura tends to acts as a barrier separating the tumor from the optic apparatus and hypothalamus, even if there is suprasellar extension of the tumor (53). Tumors located in an enlarged sella may be amenable to trans-sphenoidal surgery (30,54). Sellar craniopharyngiomas typically present with a higher rate of pituitary dysfunction, which only rarely improves after trans-sphenoidal surgery (29). Recent reports suggest that the risk of endocrine dysfunction is not higher after trans-sphenoidal approach, and recommend that this approach be taken regardless of the preoperative endocrine status. With respect to the trans-sphenoidal technique, a few important points need to be noted. A wide exposure of the sella, with the margins extending laterally to the cavernous sinuses, is necessary to maximize visualization of the anatomy of the area. Opening of the dura needs to be done carefully in order not to breach the tumor capsule or to damage the normal pituitary gland which is frequently found in a ventral sellar location (53). The gland, once visualized, must be lateralized to reach the tumor. Capsular attachments to the dura or the stalk may hamper the removal of the tumor while ensuring the integrity of these structures, and in some cases it may be necessary to sharply cut the stalk to remove the capsule. CSF leak will occur when incising the diaphragm to remove intrasellar portions of the tumor, or if there are dural defects or absence of the diaphragma sellae. CSF leaks can be minimized through the preoperative placement of a lumbar drain and by packing the sella after tumor resection, with fat or muscle. Trans-sphenoidal resection can be carried out safely in the pediatric population, even if the sphenoid sinuses are not pneumatized using appropriate guidance strategies to ensure that the surgeon is staying in the midline. Transcallosal or transcortical approaches may be required for craniopharyngiomas that have large intraventricular components, or as part of combined approaches in dealing
with craniopharyngiomas with more complex morphology (49). Other neurosurgical approaches include a subtemporal approach which has been used for retrochiasmatic tumors with an inferior extension, or for those deviating from midline (55). Modern innovations in minimally invasive neurosurgery have made available the use of neuroendoscopes, and experience in this area is gaining momentum (56,57). In craniopharyngioma surgery also, endoscopic treatment has been attempted with encouraging early results. Endoscopy can be used as either an adjunct for trans-sphenoidal surgery or for intraventricular approaches. In the latter, the presence of hydrocephalus greatly facilitates the use of the endoscope. However, strong data to support this approach have yet to be published. For predominantly cystic craniopharyngiomas, intracavitary therapy remains an option (Fig. 6). This form of
Figure 6 Axial CT of 1-year-old female with cystic craniopharyngioma showing cystic suprasellar lesion with Ommaya reservoir in place. This reservoir system was used to deliver bleomycin sclerotherapy over several cycles.
Chapter 40: Craniopharyngioma: Neurosurgical Management
treatment may be as limited as inserting a catheter into the cyst and placing an Ommaya reservoir for repeated cyst aspirations over time, to the instillation of various radio-isotopes, or sclerosing agents into the cyst. Agents which have been utilized in the past include 32 P, yttrium (90 Y), or 125 I. 32 P has an effective range of radiation less than 1 mm while yttrium has a shorter half-life (58). Many centers have reported on the use of intracystic bleomycin with variable success rates (59,60). (Kim, 2007) (61).
Adjuvant Therapy Radiation Therapy The high morbidity associated with the resection of some craniopharyngiomas has led to the use of radiation treatment (62,63), and this has now become an important adjunct in the management protocol of patients with these tumors. Since craniopharyngiomas are radiosensitive, the use of radiation is especially indicated when only subtotal or partial tumor resection has been achieved. Although there are many reports on the improved survival with the use of radiation, there are no detailed studies comparing the different types of radiation protocols and their effects. We highlight below several of the radiation therapy strategies used for craniopharyngioma.
External Beam Radiation External beam radiation is designed to irradiate the target with a high dose of radiation, while at the same time increasing normal tissue tolerance by fractionation of the total dose. The present recommended treatment is using a megavoltage photon beam, delivering a dose of 5400 cGy in fractions of 180 cGy (64). This mode of treatment is the one used most frequently for craniopharyngiomas. Patients who have undergone subtotal resection or biopsy alone should be considered for this mode of therapy. Patients fare better when they undergo radiation earlier in time relative to their initial intervention, rather than when the tumor recurs.
Stereotactic Radiosurgery—Gamma Knife Gamma knife uses convergence of multiple beams to achieve a highly focused, localized radiation field, such that a single dose will deliver the desired radiation to the area of interest. Although the dose delivered beyond the confines of the chosen target is minimal, the important structures in the area, particularly the chiasm appear to be vulnerable. Calculations of radiation delivery must be made in such a way that the chiasm does not receive doses much higher than 10 cGy (65), while delivering a maximal dose of 1900 to 3200 cGy to the tumor itself (66). Thus, if a tumor is in too close proximity to the optic pathway, Gamma knife radiosurgery is not recommended. Furthermore, treatment of tumors that have a predominant cystic component can lead to cyst enlargement, requiring subsequent intervention (67). Gamma knife radiosurgery of craniopharyngiomas offers promising results; however, the current studies do not have a long follow-up and further analysis is clearly indicated.
Chemotherapy Chemotherapy is an additional adjuvant option in recurrent disease when radiotherapy is not feasible, either in patients that are very young, or when the brain has already been irradiated. Chemotherapy can be in different forms, including systemic injection of Adriamycin (doxorubicin) and CCNU (lomustin) (68). Vincristine, BCNU (carmustine), and procarbazine treatment has also been reported (69), with minimal side effects (69). There are currently no large series looking at
579
the effectiveness of chemotherapy in the treatment of craniopharyngiomas.
OUTCOME AND PROGNOSIS Complete surgical resection has been shown repeatedly to correlate with disease cure (13,37,46,49). Few would argue that complete surgical resection is the preferred mode of management provided that patient morbidity is acceptable. The problem, however, is achieving a gross total resection with minimal morbidity especially in patients with the large, retrochiasmatic tumors. The strongest predictor of recurrence is residual tumor at the time of operation or on postoperative imaging (37,70). Adamantinomatous craniopharyngioma has a greater tendency to recur. This is likely related to the fact that it is more often invasive of the pia (11,49). In terms of molecular markers of recurrence or aggressive behavior there are several case reports and series of low power. Only statistically significant results will be discussed here. It has been suggested that a high ratio of retinoic acid receptor gamma (RARG) to retinoic acid receptor beta (RARB) predicts a tendency to recur (71). Further to this, the same group has demonstrated a strong correlation of RARG to Cathepsin K and RARB to Cathepsin D (72). The significance of this relationship to recurrence remains under investigation. Immunoreactivity to p53 and presence of whorl-like arrays on histopathology are associated to recurrence/regrowth of craniopharyngioma (70). Alpha(2beta1) integrin expressed by craniopharyngiomas as well as vitronectin expressed in surrounding tissue is associated with a higher rate of recurrence (73). Once thought to represent an increased risk of recurrence, VEGF expression does not predict outcome (74).
CONCLUSION Craniopharyngioma is a complex, centrally situated intracranial tumor that has evoked much debate and controversy over the years especially insofar as patient treatment is concerned. Certain craniopharyngiomas, such as those that are prechiasmatic or intrasellar in location, are best managed by aggressive neurosurgical resection with the goal of achieving a gross total resection. Complex, large retrochiasmatic craniopharyngiomas with significant hypothalamic infiltration should be approached cautiously by neurosurgeons. In these patients, a maximal safe neurosurgical resection should be employed followed by adjuvant radiation therapy strategies. The implementation of conformal radiation therapy techniques has improved the delivery of radiation therapy to the tumor while sparing regions of normal brain as much as possible. Cystic craniopharyngiomas will require cyst-directed therapies. In the next decade, the role of neuroendoscopic resection of craniopharyngiomas will be defined.
REFERENCES 1. Bunin GR, Surawicz TS, Witman PA, et al. The descriptive epidemiology of craniopharyngioma. [see comment]. J Neurosurg. 1998;89:547–551. 2. Haupt R, Magnani C, Pavanello M, et al. Epidemiological aspects of craniopharyngioma. J Pediat Endocrinol. 2006;1:289–293. 3. Kuratsu J, Ushio Y. Epidemiological study of primary intracranial tumors in childhood. A population-based survey in Kumamoto Prefecture, Japan. Pediat Neurosurg. 1996;25:240–246.
580
Cook and Rutka
4. Carmel PW, Antunes JL, Chang CH. Craniopharyngiomas in children. Neurosurgery. 1982;11:382–389. 5. Banna M. Craniopharyngioma in adults. Surg Neurol. 1973;1:202–204. 6. Banna M, Hoare RD, Stanley P, et al. Craniopharyngioma in children. J Pediat. 1973;83:781–785. 7. Hopper N, Albanese A, Ghirardello S, et al. The pre-operative endocrine assessment of craniopharyngiomas. J Pediat Endocrinol. 2006;1:325–327. 8. DeVile CJ, Grant DB, Hayward RD, et al. Growth and endocrine sequelae of craniopharyngioma. Arch Dis Childhood. 1996;75:108–114. 9. Di Battista E, Naselli A, Queirolo S, et al. Endocrine and growth features in childhood craniopharyngioma: A mono-institutional study. J Pediat Endocrinol. 2006;1:431–437. 10. Stark KL, Kaufman B, Lee BC, et al. Visual recovery after a year of craniopharyngioma-related amaurosis: Report of a nine-yearold child and a review of pathophysiologic mechanisms. J Aapos: Am Assoc Pediat Ophthalmol Strab. 1999;3:366–371. 11. Pierre-Kahn A, Recassens C, Pinto G, et al. Social and psychointellectual outcome following radical removal of craniopharyngiomas in childhood. A prospective series. Childs Nerv Syst. 2005;21:817–824. 12. Roth C, Wilken B, Hanefeld F, et al. Hyperphagia in children with craniopharyngioma is associated with hyperleptinaemia and a failure in the downregulation of appetite. Eur J Endocrinol. 1998;138:89–91. 13. Zuccaro G. Radical resection of craniopharyngioma. Childs Nerv Syst. 2005;21:679–690. 14. Satoh H, Uozumi T, Arita K, et al. Spontaneous rupture of craniopharyngioma cysts. A report of five cases and review of the literature. Surg Neurol. 1993;40:414–419. 15. Kulkarni V, Daniel RT, Pranatartiharan R. Spontaneous intraventricular rupture of craniopharyngioma cyst. Surg Neurol. 2000;54:249–253. 16. Rathke H. Uber die Entstehung der Glandula pituitaria. Arch Anat Physiol Wissen Med. 1838;5:482–485. 17. Mott FW, Barrett JO. Three cases of tumors of third ventricle. Arch Neurol Psychiatry. 1899;1:417–440. 18. Luschka H. Der Hirnangang und die Steissdrtise des Menschen. Berlin: Reimer, 1860. 19. Cushing H. Intracranial Tumors. Notes upon a Series of Two Thousand Verified Cases with Surgical Mortality Percentages Pertaining Thereto. Springfield, IL: Thomas, 1932. 20. Yamada H, Haratake J, Narasaki T, et al. Embryonal craniopharyngioma. Case report of the morphogenesis of a craniopharyngioma. Cancer. 1995;75:2971–2977. 21. Hanna E, Weissman J, Janecka IP. Sphenoclival Rathke’s cleft cysts: Embryology, clinical appearance and management. Ear Nose Throat J. 1998;77:396–399. 22. Critchley M, Ironside RN. The pituitary adamantinomata. Brain. 1926;49:437–481. 23. Ingraham FD, Scott HW. Craniopharyngiomas in children. J Pediat. 1946;29:95–116. ¨ ¨ 24. Erdheim J. Uber Hypophysenganggeschwulste und Hirncholesteatome. Sitzber Kais Akad Wiss, Math-natwiss Cl Abt. 1904;3:537–726. 25. Goldberg GM, Eshbaugh DE. Squamous cell nests of the pituitary gland as related to the origin of craniopharyngiomas. A study of their presence in the newborn and infants up to age four. Arch Pathol Lab Med. 1960;70:293–299. 26. Carmichael HT. Squamous epithelial rests in the hypophysis cerebri. Arch Neurol Psychiatry. 1931;26:966–975. 27. Luse SA, Kernohan JW. Squamous-cell nests of the pituitary gland. Cancer. 1955;8:623–628. 28. Miller DC. Pathology of craniopharyngiomas: Clinical import of pathological findings. Pediat Neurosurg. 1994;1:11–17. 29. Al-Brahim NY, Asa SL. My approach to pathology of the pituitary gland. J Clin Pathol. 2006;59:1245–1253. 30. Jagannathan J, Dumont AS, Jane JA, Jr. Diagnosis and management of pediatric sellar lesions. Front Horm Res. 2006;34:83– 104.
31. Rienstein S, Adams EF, Pilzer D, et al. Comparative genomic hybridization analysis of craniopharyngiomas. J Neurosurg. 2003;98:162–164. 32. Sarubi JC, Bei H, Adams EF, et al. Clonal composition of human adamantinomatous craniopharyngiomas and somatic mutation analyses of the patched (PTCH), Gsalpha and Gi2alpha genes. Neurosci Lett. 2001;310:5–8. 33. Gorski GK, McMorrow LE, Donaldson MH, et al. Multiple chromosomal abnormalities in a case of craniopharyngioma. Cancer Genet Cytogenet. 1992;60:212–213. 34. Karnes PS, Tran TN, Cui MY, et al. Cytogenetic analysis of 39 pediatric central nervous system tumors. Cancer Genet Cytogenet. 1992;59:12–19. 35. Rickert CH, Paulus W. Lack of chromosomal imbalances in adamantinomatous and papillary craniopharyngiomas. J Neurol Neurosurg Psychiat. 2003;74:260–261. 36. Yoshimoto M, de Toledo SR, da Silva NS, et al. Comparative genomic hybridization analysis of pediatric adamantinomatous craniopharyngiomas and a review of the literature. J Neurosurg. 2004;101:85–90. 37. Hoffman HJ. Surgical management of craniopharyngioma. Pediat Neurosurg. 1994;1:44–49. 38. Rutka JT, Hoffman HJ, Drake JM, et al. Suprasellar and sellar tumors in childhood and adolescence. Neurosurg Clin North Am. 1992;3:803–820. 39. Wang KC, Hong SH, Kim SK, et al. Origin of craniopharyngiomas: Implication on the growth pattern. Childs Nerv Syst. 2005;21:628–634. 40. Van Effenterre R, Boch AL. Craniopharyngioma in adults and children: A study of 122 surgical cases. J Neurosurg. 2002;97:3– 11. 41. Karavitaki N, Cudlip S, Adams CB, et al. Craniopharyngiomas. Endocrine Rev. 2006;27:371–397. 42. Hald JK, Eldevik OP, Skalpe IO. Craniopharyngioma identification by CT and MR imaging at 1.5 T. Acta Radiol. 1995;36:142– 147. 43. Harwood-Nash DC. Neuroimaging of childhood craniopharyngioma. Pediat Neurosurg. 1994;1:2–10. 44. Rossi A, Cama A, Consales A, et al. Neuroimaging of pediatric craniopharyngiomas: A pictorial essay. J Pediat Endocrinol. 2006;1:299–319. 45. Curran JG, O’Connor E. Imaging of craniopharyngioma. Childs Nerv Syst. 2005;21:635–639. 46. Hoffman HJ, De Silva M, Humphreys RP, et al. Aggressive surgical management of craniopharyngiomas in children. J Neurosurg. 1992;76:47–52. 47. Fahlbusch R, Honegger J, Paulus W, et al. Surgical treatment of craniopharyngiomas: Experience with 168 patients. J Neurosurg. 1999;90:237–250. 48. Tomita T, McLone DG. Radical resections of childhood craniopharyngiomas. Pediat Neurosurg. 1993;19:6–14. 49. Yasargil MG, Curcic M, Kis M, et al. Total removal of craniopharyngiomas. Approaches and long-term results in 144 patients. J Neurosurg. 1990;73:3–11. 50. Sklar CA. Craniopharyngioma: Endocrine abnormalities at presentation. Pediat Neurosurg. 1994;1:18–20. 51. Honegger J, Buchfelder M, Fahlbusch R. Surgical treatment of craniopharyngiomas: Endocrinological results. J Neurosurg. 1999;90:251–257. 52. Baskin DS, Wilson CB. Surgical management of craniopharyngiomas. A review of 74 cases. J Neurosur. 1986;65:22–27. 53. Laws ER Jr. Transsphenoidal removal of craniopharyngioma. Pediat Neurosurg. 1994;1:57–63. 54. Norris JS, Pavaresh M, Afshar F. Primary transsphenoidal microsurgery in the treatment of craniopharyngiomas. Br J Neurosurg. 1998;12:305–312. 55. Symon L, Pell MF, Habib AH. Radical excision of craniopharyngioma by the temporal route: A review of 50 patients. Br J Neurosurg. 1991;5:539–549. 56. de Divitiis E, Cappabianca P, Gangemi M, et al. The role of the endoscopic transsphenoidal approach in pediatric neurosurgery. Childs Nerv Syst. 2000;16:692–696.
Chapter 40: Craniopharyngioma: Neurosurgical Management 57. Teo C. Application of endoscopy to the surgical management of craniopharyngiomas. Childs Nerv Syst. 2005;21:696–700. 58. Backlund EO. Treatment of craniopharyngiomas: The multimodality approach. Pediat Neurosur. 1994;1:82–89. 59. Broggi G, Franzini A. Bleomycin for cystic craniopharyngioma. [comment]. J Neurosurg. 1996;84:1080–1081. 60. Mottolese C, Stan H, Hermier M, et al. Intracystic chemotherapy with bleomycin in the treatment of craniopharyngiomas. Childs Nerv Syst. 2001;17:724–730. 61. Hukin J, Steinbok P, Lafay-Cousin L, et al. Intracystic bleomycin therapy for craniopharyngioma in children: The Canadian experience. Cancer. 2007;109;2124–2131. 62. Kramer S, McKissock W, Concannon JP. Craniopharyngiomas. Treatment by combined surgery and radiation therapy. J Neurosurg. 1961;18:217–226. 63. Kalapurakal JA. Radiation therapy in the management of pediatric craniopharyngiomas—a review. Childs Nerv Syst. 2005;21:808–816. 64. Wara WM, Sneed PK, Larson DA. The role of radiation therapy in the treatment of craniopharyngioma. Pediat Neurosurg. 1994;1:98–100. 65. Lunsford LD, Pollock BE, Kondziolka DS, et al. Stereotactic options in the management of craniopharyngioma. Pediat Neurosurg. 1994;1:90–97. 66. Chung WY, Pan HC, Guo WY, et al. Protection of visual pathway in gamma knife radiosurgery for craniopharyngiomas. Stereotact Funct Neurosurg. 1998;1:139–151. 67. Chung WY, Pan DH, Shiau CY, et al. Gamma knife radiosurgery for craniopharyngiomas. J Neurosurg. 2000;3:47–56.
581
68. Lippens RJ, Rotteveel JJ, Otten BJ, et al. Chemotherapy with Adriamycin (doxorubicin) and CCNU (lomustine) in four children with recurrent craniopharyngioma. Eur J Paediat Neurol. 1998;2:263–268. 69. Bremer AM, Nguyen TQ, Balsys R. Therapeutic benefits of combination chemotherapy with vincristine, BCNU, and procarbazine on recurrent cystic craniopharyngioma. A case report. J Neuro Oncol. 1984;2:47–51. 70. Tena-Suck ML, Salinas-Lara C, Arce-Arellano RI, et al. Clinicopathological and immunohistochemical characteristics associated to recurrence/regrowth of craniopharyngiomas. Clin Neurol Neurosurg. 2006;108:661–669. 71. Lefranc F, Chevalier C, Vinchon M, et al. Characterization of the levels of expression of retinoic acid receptors, galectin-3, macrophage migration inhibiting factor, and p53 in 51 adamantinomatous craniopharyngiomas. J Neurosurg. 2003;98:145–153. 72. Lubansu A, Ruchoux MM, Brotchi J, et al. Cathepsin B, D and K expression in adamantinomatous craniopharyngiomas relates to their levels of differentiation as determined by the patterns of retinoic acid receptor expression. Histopathology. 2003;43:563– 572. 73. Lefranc F, Mijatovic T, Decaestecker C, et al. Monitoring the expression profiles of integrins and adhesion/growth-regulatory galectins in adamantinomatous craniopharyngiomas: Their ability to regulate tumor adhesiveness to surrounding tissue and their contribution to prognosis. Neurosurgery. 2005;56:763–776. 74. Xu J, Zhang S, You C, et al. Microvascular density and vascular endothelial growth factor have little correlation with prognosis of craniopharyngioma. Surg Neurol. 2006;66:S30–S34.
41 Epidermoids, Dermoids, and Other Cysts of the Skull Base Samuel P. Gubbels and Bruce J. Gantz
Multiple epidermoid tumors almost never occur and no familial predisposition to the development of these lesions has been reported. Recently, intracranial epidermoid cysts have been reported in three patients with craniovertebral junction anomalies (20); the first known association of these lesions with other congenital malformations.
INTRODUCTION Cystic lesions of the skull base, in general, represent growth of normally occurring tissues in abnormal or aberrant locations, rather than true neoplasia. Though skull base cysts are uncommon lesions, the difficulty in obtaining complete removal and the potential morbidity incurred in their treatment make patients afflicted with them quite memorable to the neurotologist and/or neurosurgeon involved in their care. This chapter reviews the clinical and pathological findings of the most common cysts affecting the skull base and discusses the diagnosis and treatment of these lesions.
Embryology Von Remak in 1854 initially proposed that epidermoids occurred due to an error in embryological development leading to entrapment of keratinizing ectodermal elements with subsequent growth of the aberrantly located tissue. The timing of this embryological event is postulated to have occurred at the time of closure of the neural groove, between the third and fifth week of development (21–24).Von Remak’s theory continues to be the most widely accepted theory for the formation of these lesions, though the timing of the embryological error and subtype (neuro- vs. cutaneous ectoderm) of the entrapped tissue have been the source of debate through the decades (25–27). Other theories regarding the origin of epidermoids have been proposed including Virchow’s theory of squamous metaplasia (4) and Fleming and Botterel’s multipotential embryonic cell rest theory (28). More recent publications have proposed a modification of Fleming and Botterel’s theory whereby multipotential cells are carried from a medial to lateral position along with migrating otic capsular elements during embryogenesis (29–31). Kountakis demonstrated migratory properties of CPA epidermoid cells in vitro similar to those of acquired cholesteatomas and, because these properties are unique to epithelium of the first branchial groove, concluded that epidermoids originate from the first branchial groove (32). Though this theory does not rule out theories such as Virchow’s or Fleming and Botterel’s, it does indirectly support the concept that entrapment of would-be first branchial cleft epithelium during embryogenesis ultimately leads to epidermoid formation.
EPIDERMOIDS Incidence and Epidemiology In 1829, the French pathologist Cruveilhier was the first to offer a systematic description of a series of epidermoids, which he termed “pearly tumors” (1). Cruveilhier reported the incidental finding at autopsy of a large epidermoid tumor in a man who had died of a head injury. In describing this case, along with two cases previously reported by Dumeril and Le Prestre (2), Cruveilhier commented on the notable size and extent that epidermoid tumors may attain without producing symptoms (1). These lesions were known for the next 100 years as “Cruveilhier’s pearly tumors” until Muller in 1838 first used the term cholesteatoma to describe them after noting the presence of cholesterol crystals within the matrix of the lesions (3). The prolific German pathologist Rudolf Virchow in 1855 concluded that pearly tumor was the more correct term to be used in the description of these lesions as he found the presence of cholesterol crystals to be inconsistent (4). Bostroem in 1897 coined the term epidermoids to describe these lesions, the term used most commonly today (5). Epidermoids represent 0.2 to 1.5% of all intracranial tumors and 6 to 14% of all tumors of the cerebellopontine angle (CPA) (6–8) where they are the third most common lesion, following schwannomas and meningiomas. Approximately 30 to 60% of all epidermoids occur in the CPA (6,9–11) followed in order of decreasing frequency by the parasellar region, paraclival area, lateral recess of 4th ventricle, and the petrous apex (12,13). In addition to the intracranial locations above, epidermoids can occur in the diploe of the calvarium (intraosseous) where they can present as a painless swelling with an associated calvarial defect (14). Intraosseous epidermoids represent up to 25% of cases in two larger series and are to be differentiated from the intracranial type of these lesions (15,16). The age of incidence of intracranial epidermoids is from birth to 80 years with the majority identified by the third or fourth decades of life (6,17). A male preponderance has been reported with a male to female ration of 5:4 (9,18,19).
Pathology Epidermoids have a characteristic appearance making them easily recognized on gross inspection. The external surface is silky with a white-gold, mother-of-pearl appearance and sheen, often with multiple delicate vessels evident on the surface. The capsule can be lobulated or smooth and tears easily with application of a shearing force. The tumor is malleable and compresses easily due to the caseous core of squamous debris. Section of the cyst reveals a capsule consisting of a thin layer of stratified, keratinizing squamous epithelium often with multiple foci of calcification surrounded by a thin layer of fibrous soft tissue. The cyst is filled with desquamated keratin debris and cholesterol crystals with a soft, waxy texture. The keratin debris within the core has a lamellar or onion-skin gross appearance (6,9,33) and is essentially 583
584
Gubbels and Gantz
avascular (34). Some epidermoids have a high triglyceride content in addition to the presence of cholesterol crystals (35). The growth rate of epidermoids is linear and to be differentiated from the exponential pattern seen with a neoplasm, either benign or malignant (24). It is only the basal layers of an epidermoid that undergo cell division, followed by progressive maturation and ultimately death of the overlying layers of the epithelium which eventually slough into the central core of the lesion. With time the central, nonviable core represents the bulk of the epidermoid with the viable cells displaced circumferentially toward the periphery of the lesion. Because of this growth pattern and the slow rate of expansion, epidermoids characteristically envelop surrounding nerves, vessels, and other critical structures rather than displacing them like most benign neoplasms. After filling the intracranial subarachnoid space from which it originated, the epidermoid then extends to adjacent spaces, eroding bone in the process (31). The capsule of the cyst typically will insinuate itself into surrounding structures as it erodes the bone in the area, conforming to anatomical features in the area as it expands. In addition to engulfing cranial nerves and vessels in the area, epidermoids can cause atrophy, ischemic injury, and paresis due to the interaction with the cerebral and cerebellar parenchyma (36). This pattern of growth combined with the relative friability of the epithelial layer make complete extirpation of these lesions difficult and, not infrequently, impossible without the sacrifice of the involved vessels and nerves, which is to be avoided.
Clinical Manifestations Due to their slow rate and pattern of growth, epidermoids can be asymptomatic and found incidentally on imaging performed for other reasons. More commonly epidermoids will present with a long, protracted course of sometimes mild and vague clinical manifestations. Early series of patients with epidermoids reported some patients with duration of symptoms attributable to the lesion as long as 53 years (34). Modern imaging has enabled earlier identification of these lesions with some reports of duration of symptoms similar to those reported in series of acoustic neuromas (11,16,37–39). Despite this, some modern series report delays in diagnosis of 20 years or more (11,13,30,37). Epidermoids present with a spectrum of symptoms similar to those seen with an acoustic neuroma or other expansile lesion of the CPA, namely cranial nerve deficits, which occur in 80 to 90% of patients (30,31). The eighth cranial nerve is most commonly involved (40–93%) (16,30,31,37,39–42) in most series of intracranial epidermoids, and may manifest as hearing loss, vertigo, tinnitus, disequilibrium or gait disturbance. Some have found the facial nerve to be more frequently affected on presentation (36). In contrast to the stretch injury to the facial nerve that can occur in cases of acoustic neuroma, epidermoids are thought to engulf the facial nerve and cause axonotmesis or neurotmesis due to the resulting ischemic injury (43), generally manifesting as facial weakness or hemifacial spasm. Trigeminal nerve involvement, producing symptoms of facial pain, numbness, corneal reflex abnormalities, and masticator muscle weakness, is more common on presentation (25–52%) in epidermoids than acoustic neuromas (13,16,30,31,37,39,41). The facial pain seen in cases of epidermoid tumors is atypical for trigeminal neuralgia in that it is longer in duration and may not be accompanied by sensory or motor dysfunction of the fifth nerve (44). Three mechanisms have been proposed to account for the trigeminal neuralgia seen with epidermoids: direct compression of the nerve root (6), indirect compression of the nerve due to vessel displace-
Figure 1 Coronal T1-weighted MRI with gadolinium of right petrous epidermoid.
ment from the enlarging cyst and toxic neuritis as a result of cyst content leakage (37,45). Larger epidermoids may involve cranial nerves III, IV, VI, IX, X, and XI to produce visual deficits, diplopia, hoarseness, and dysphagia. Headache occurs frequently with epidermoids (30,31,36,37) and may be due to the expansile nature of the lesion (which may also produce papilledema secondary to increased intracranial pressure) or to a toxic effect from leakage of the cyst contents. Seizures due to epidermoids have been described but are an uncommon manifestation of the disease (46).
Radiology Please refer to chapter 4 for a detailed discussion of the imaging characteristics of epidermoid cysts of the skull base. Figures 1–6 show representative imaging of a skull base epidermoid.
Figure 2 Axial T1-weighted MRI with gadolinium of the same lesion as Figure 1.
Chapter 41: Epidermoids, Dermoids, and Other Cysts of the Skull Base
Figure 3 Axial constructive interference in steady state (CISS) image of same patient as Figures 1 and 2.
585
Figure 6 Axial bone window CT scan of right petrous epidermoid. Compare with Figure 2.
Treatment
Figure 4 Axial diffusion-weighted MRI image of the same petrous epidermoid as demonstrated in Figures 1–3.
Epidermoids of the skull base are best treated with surgical excision. Chemotherapy has no role in the treatment of these lesions as they are not true neoplasms. The use of external beam radiation therapy has been reported in one case of a recurrent epidermoid (47) but should be reserved for symptomatic treatment in patients who are not surgical candidates, if at all. In general, complete cyst removal should be pursued in all cases but, due to the infiltrative nature of these lesions, may not be possible without incurring significant morbidity and even mortality. Epidermoids engulf and become intimately involved with surrounding neurovascular structures making complete excision difficult if not impossible without risking potentially devastating consequences from sacrifice of cranial nerves, vascular injury, or damage to the cerebellum, brainstem, or temporal lobe. Incomplete excision, leaving cyst matrix on cranial nerves and vasculature intimately involved with the lesion, is an acceptable alternative in some cases as “recurrence may occur slowly and reoperation may not be required for many years” (7,37,44). The rates of total cyst excision vary considerably in published series with rates as low a 0% (11,36) and as high as 80 to 97% (30,31). Second operations are required in up to one-third of patients (44), generally many years after initial resection. Multiple approaches have been employed to access and excise these lesions including suboccipital, translabyrinthine, transotic, transcochlear, subtemporal, and petrosal routes (31,36,37,48,49). Please refer to table of contents of this text to be divected to chapter with detailed discussions of these surgical approaches.
Outcome and Prognosis
Figure 5 Coronal bone window CT scan of a right petrous epidermoid. Compare with Figure 1.
The rates of total cyst removal vary from 0% to 97% in modern published series with rates of recurrence requiring a second operation ranging from 0% to 36%. Overall published recurrence rates range from 0% to 55% of patients and the reported time to recurrence of epidermoids ranges from 36 to 264 months, highlighting the importance of long-term follow-up with periodic MRI surveillance in all patients with epidermoids, whether or not total excision was achieved
586
Gubbels and Gantz
at the initial surgery (44). Mortality rates in modern series ranges from 0% to 16%, a significant improvement from the premicrosurgical era (44). Complications of surgery for epidermoids of the skull base are similar to those seen in the treatment of other CPA neoplasms including CSF leak, cranial nerve injury, infection, aseptic meningitis, hydrocephalus, pulmonary embolus, headache, seizures, and aspiration (31,37,49). Cranial nerve eight is the most frequently injured during excision of these lesions, though many series employ a translabyrinthine possibly transcochlear approach to address these lesions which may skew the interpretation of the rates of postoperative eighth nerve dysfunction. It is difficult to assess the rates of eighth nerve dysfunction postoperatively from published series due to variability in the reporting of pre-existing dysfunction, lack of mention of whether hearing preservation was attempted and absence of clear descriptions of the nature of the postoperative dysfunction (vestibular vs. auditory, conductive vs. sensorineural hearing loss, means of auditory assessment). Nevertheless, reported rates of hearing preservation (as loosely defined) when attempted range from 15% to 72% (17,31,37,49). Similarly, interpretation of the rates of seventh cranial nerve dysfunction postoperatively is difficult due to variability in reporting but range from an 8% to 50% incidence of some level of facial nerve dysfunction (17,31,37,39). It is important to note that most series include patients with some degree of preoperative facial nerve dysfunction or who had the facial nerve mobilized to enable complete cyst removal, both of which would obviously bias the reported rates of postoperative dysfunction. In general, the rates of postoperative cranial nerve dysfunction after removal of intracranial epidermoids are highly variable and potentially subject to a number of factors including age, coexistent medical problems, level of preoperative dysfunction, extent of nerve involvement by the epidermoid, meticulousness of cyst dissection and tissue handling, level of scrutiny in the evaluation of postoperative dysfunction, and administration of corticosteroids. Furthermore, the risk of injury to the facial and other involved cranial nerves is higher than for acoustic tumors of similar size given the tendency of epidermoids to encircle the nerves (46). Intraoperative neurophysiological monitoring of involved cranial nerves, especially the facial nerve, is recommended (50). When assessing the outcomes of surgical interventions for epidermoids of the skull base, it is important to appreciate the natural history of these lesions with their associated morbidity and mortality. Reported complications associated with untreated intracranial epidermoids include hearing loss, facial weakness, facial numbness, lower cranial neuropathies, diplopia, blindness, headaches, cerebellar dysfunction, encephalitis, seizures, hydrocephalus, and death (31,36,46). Recurrent aseptic meningitis occurring spontaneously in the setting of an epidermoid has been described in a number of reports, some as the initial presentation of the lesion (44). In addition, aseptic meningitis occurring after excision of epidermoid tumors has been reported in many series, thought secondary to a toxic effect of spilled keratin debris within the subarachnoid space (30,31,36,37). In the evaluation and management of patients with epidermoid cysts one must maintain vigilance for the development of aseptic meningitis and initiate treatment with corticosteroids and possibly antibiotics if suspected. A rare but feared complication of intracranial epidermoids is the development of squamous cell carcinoma, either a de novo lesion discovered at the first operation for epidermoid (31,51–57), incidentally at autopsy (58–60) or,
more frequently, in the setting of an incompletely resected lesion. Most reports describe a delay of 3 months to 33 years (61) before discovery of carcinoma in the setting of an incompletely excised lesion (22,53,62–67). There appears to be a male predominance and patients typically had rapidly progressive and more severe symptoms than those seen in benign epidermoids. Some patients had an episode of aseptic meningitis prior to the development of carcinoma within the epidermoid, supporting the concept that chronic inflammation of the cyst matrix ultimately leads to malignant transformation (22), analogous to the pattern seen when carcinoma develops within a burn scar or area of chronic ulceration. Enhancement of a portion of an incompletely resected epidermoid tumor on CT or MRI, especially in the presence of atypical or severe symptoms, should alert one to the possibility of carcinoma (31,68). Improved clinical outcomes have been described with the use of adjuvant treatment with radiation (52–55,57,64,65,69–72). Asahi described an improvement in mean survival from 4 months to 15 months with the addition of external beam radiation therapy to surgical resection of carcinoma arising within an intracranial epidermoid (61). Tamura reviewed the use of stereotactic radiotherapy in the setting of carcinoma arising within an intracranial epidermoid and found median survival times of 1, 18, and 44 months with the use of surgery alone, surgery plus external beam radiation, and surgery plus stereotactic radiotherapy respectively, differences that the authors found to be statistically significant (73). Murase stressed the importance of combination chemotherapy and radiation in the treatment of these lesions (69), though the benefit of adding chemotherapy in these cases remains unproven (61). Despite these measures the prognosis overall for carcinoma in the setting of an epidermoid of the skull base remains poor (31,46,61).
DERMOIDS Incidence and Epidemiology Dermoids, like epidermoids, are cysts lined by stratified, keratinizing squamous epithelium but differ from epidermoids in that they also contain mesodermal elements such as hair, sebaceous glands, sweat glands or, rarely, teeth, bone, or cartilage (44,74,75). Dermoids are to be differentiated from teratomas, which are true neoplasms that originate from a misplaced rest of embryonic germ cells which progress to form a tumor composed of well-differentiated tissue derivatives of all three germ layer in an organ-like pattern (76). Like epidermoids, dermoids grow through the division of the outer lining of the cyst with progressive expansion of the core of the cyst as cellular debris accumulates. The growth pattern again is linear, not exponential as with a true neoplasm (30). Verratus in 1745 was the first to identify a dermoid on autopsy of a 40-year-old woman who died from a febrile illness (77). Masses of hair, in addition to abundant squamous debris were found in pathological examination of the lesion. In 1860 Lannelogue and Achard (78) reported the first posterior fossa dermoid occurring in a child. Six years later, Toynbee and Hinton described cystic lesions with “masses of hair” within the mastoid and middle ear space (79,80). The first reported attempt at surgical removal of an intracranial dermoid was by Horrax in 1922 but not until 1934 was a dermoid successfully removed from the posterior fossa by Tytus and Pennybacker (81,82). Intracranial dermoids are rare lesions, accounting for 0.04 to 0.7% of all intracranial neoplasms with an incidence one quarter to one half that of intracranial epidermoids (33,76,83–88). Dermoids can occur in many intracranial locations both intra- and extradurally (89), and are thought
Chapter 41: Epidermoids, Dermoids, and Other Cysts of the Skull Base
to typically occur in more midline locations when compared to epidermoids. Above the tentorium, dermoids most frequently occur in the frontobasal, suprasellar, parasellar, cavernous sinus, and temporal regions (85,90–92). Infratentorial locations for dermoids include the occiput, cerebellopontine angle, and prepontine areas. Dermoids have also been reported within the temporal bone in the mastoid complex (93), tympanum (94), and petrous apex (95). Dermoids can occur in all age groups but are found more frequently in children than epidermoids, especially when considering dermoids of the posterior fossa (96). In reviewing multiple case reports of posterior fossa dermoids one study found a median age of 2 years (range 6 months to 27 years) with relatively equal sex distribution (89). Dermoids of the posterior fossa have been associated with Klippel–Feil syndrome in a number of reports (97–104). There are three case reports in the literature of dermoids in children with Goldenhar syndrome, though this association is less clear (105–107). No familial predilection to the development of intracranial dermoids has been reported.
Embryology Similar embryological events are thought to lead to the formation of dermoids and epidermoids (see above). Some have postulated that dermoids form as a result of the aberrant adherence of primitive mesodermal cells to developing intracranial veins (108). The predilection of dermoids for a midline position has been theorized to occur when cutaneous ectoderm is drawn intracranially as the falx cerebri or tentorium are forming from the fusion of two leaflets of dura (89).
Pathology Dermoid cysts are similar in appearance to epidermoids but have a variable number of hairs or dermal appendages (sebaceous or sweat glands) present. Fat is often present within dermoid cysts in addition to desquamated debris. Teeth, bone, cartilage, salivary glands, nerves, and lymph nodes (109) have been reported in dermoids but are rare (76). Dermoids have a variable growth rate and size upon presentation, with some reports of large posterior fossa masses in children (96) and others of small, limited tympanic lesions in adults (110). Dermoids, especially those occurring in the posterior fossa, can have an associated dermal sinus that can be complete or incomplete in its connection. Logue and Till in 1952 proposed a classification based on the position of the cyst relative to the dura and connection to an associated occipital dermal sinus as follows: (1) extradural cyst with complete dermal sinus; (2) intradural cyst without an associated sinus; (3) intradural cyst with an incomplete dermal sinus; (4) intradural dermoid cyst with a complete dermal sinus (111). In addition, dermoids may be either extra axial or intra axial in location, the most common intra axial sites being the fourth ventricle or cerebellar vermis (89). Lunardi found that the mean age of children with dermoid cyst of the posterior fossa with an associated dermal sinus was 2.4 years whereas dermoids without a dermal sinus tended to be discovered in older children (9 years average), reflecting that the presence of a dermal sinus will often lead to an earlier diagnosis of the intracranial component (96). In addition to intradural and extradural locations, dermoids have been classified as interdural when they are located between dural layers within the cavernous sinus, an uncommon location generally presenting with an oculomotor palsy (90,112,113). Rupture of dermoid and, less commonly, epidermoid cysts occurring spontaneously or during surgical removal, is a well documented complication that can result in significant morbidity. The mechanism of spontaneous rupture of
587
dermoid cysts is unclear, though Stendel hypothesized that glandular secretion under age-dependent hormonal changes leads to rapid enlargement with subsequent with rupture and spillage of fatty contents in the subarachnoid space and ventricular system (76). Rupture of dermoid cysts after head trauma (114) and during surgery (115) have also been described. Aseptic meningitis is the most frequent complication that can result from rupture of an intracranial dermoid (21) and may lead to cranial nerve fibrosis (22) or obstructive hydrocephalus due to occlusion of the ventricular outlets (115–119). Rupture of intracranial dermoids can also cause an acute (120–122) or delayed (123) cerebral vasospasm with resultant ischemia, though the pathophysiology of this process is unclear (21). The presence of fat within the subarachnoid space and ventricular system following rupture of dermoids (Fig. 10) has been well described and known to persist for years in some cases (21,118,124,125).
Clinical Manifestations Intracranial dermoids cause a spectrum of symptoms similar to those seen with epidermoids (see above). Dermoids located in the cavernous sinus can cause paresis of cranial nerves III, IV, VI, V1, and V2. Ruptured intracranial dermoids most commonly present with headaches, seizures and, when ischemia has occurred, sensory or motor defects (21). However, rupture of an intracranial epidermoid may occur without symptoms (75,124). Martinez-Lage and colleagues suggest that, at least in the case of dermoids of the posterior fossa, tumors reach a “critical size” of 3 cm where most patients develop symptoms and because of this recommend removal prior to the lesion reaching this point.
Radiology Please refer to chapter 4 for a discussion of the imaging characteristics of dermoid cysts of the skull base. Figures 7–10 show representative imaging of a left temporal dermoid.
Figure 7 Coronal T1-weighted MRI with gadolinium of a left temporal dermoid with an associated fatty component.
588
Gubbels and Gantz
Figure 8 Coronal T2-weighted MRI of the same lesion as Figure 8.
Figure 10 Sagittal T1-weighted MRI (without gadolinium) showing multiple fat lobules along thalamus and cerebellum from rupture of the dermoid seen in Figures 7–10.
Treatment Intracranial dermoid cysts are optimally treated with complete surgical removal, including the total removal of any associated dermal sinus, if present. Given the difficult anatomical relationships often encountered with these lesions complete removal is sometimes impossible. When rupture of an intracranial dermoid has occurred, thorough irrigation of the subarachnoid space is recommended (21,76). The use of a corticosteroid containing solution to irrigate the operative field has been advocated as a way to prevent the development of aseptic meningitis when rupture of a dermoid or epidermoid cyst has occurred (31,39–41,115).
Outcome and Prognosis Published reports with follow-up of 6 to 22 years have shown freedom from recurrence following removal of dermoid cysts (82,126), including subtotal resection in one patient (96). Recurrence of incompletely resected dermoids has been reported three times in the literature (45,85,89), in contrast to
Figure 9 10.
Sagittal T1-weighted MRI of the same lesion as Figures 9 and
subtotally resected epidermoids (16,33,127). Whether this reflects a lower propensity for recurrence with dermoids versus simply a lower incidence of the disease is unclear. Periodic monitoring with contrast-enhanced MRI scanning after removal of dermoids is recommended.
ARACHNOID CYSTS Incidence and Epidemiology The first report of an intracranial arachnoid cyst was by Bright in 1831 who described them as “serous cysts forming in connection with the arachnoid” (128). Maunsell in 1899 first described an arachnoid cyst in the posterior fossa (129) but not until 1932 did Mullin report the first one of these lesions in the cerebellopontine angle (130) followed by Aubry in 1937 (131). Since that point, there have been many reports of arachnoid cysts in the posterior fossa (132–143). Arachnoid cysts represent 1% of all intracranial lesions and occur most commonly in the middle fossa at the sylvian fissure (144,145). The CPA is the second most common site for the development of arachnoid cysts. Temporal bone involvement with arachnoid cysts, though well-described, is rare (140,143,144,146–152). There appears to be a male preponderance with arachnoid cysts, at least with regard to those involving the middle fossa (141,145,153). The mean age at surgery for arachnoid cysts in one review was 31 years, although 70% of patients with arachnoid cysts become symptomatic in childhood (141) and 60 to 90% of reported patients are children (140,154–156). This discrepancy is most likely a reflection of the vague symptoms that typify these lesions and the consequent delay in diagnosis and/or treatment that result from the often nonspecific symptoms. Sinha and colleagues have reported a left-sided predominance for arachnoid cysts. Arachnoid cysts have been associated with congenital anomalies such as polycystic kidney disease (157) as well as developmental syndromes such as Kabuki (158), Goldenhar (159), Chudley–McCullough (160), trisomy-12 (161), and Cri-du-chat (162). In addition, bilateral temporal arachnoid cysts have been well documented in association with glutaric aciduria type I patients (163–167), in whom the increased catabolism associated with surgical interventions may
Chapter 41: Epidermoids, Dermoids, and Other Cysts of the Skull Base
produce devastating worsening of neurological status (167). Familial arachnoid cysts have been described by multiple groups (163,165,168–170), though the pattern of inheritance remains unclear (145).
Embryology and Pathophysiology Primary (or “true”) arachnoid cysts are congenital malformations thought to form at gestational week 15 when the roof of the fourth ventricle opens through the foramina of Luschka and Magendie into the cisterna magna (143,148,150). Secondary arachnoid cysts can result from trauma, neoplasia, infection, radiation, or hemorrhage (46,151) and are thought to develop due to an adhesive arachnoiditis potentially caused by any of these insults (171). With regard to primary lesions, Starkman hypothesized that as the fourth ventricle opens into the cisterna, aberrant flow of CSF causes a splitting or duplication of the arachnoid membrane that, with further filling, results in the formation of an arachnoid cyst (172). Petrous apex arachnoid cysts are thought to arise from CSF pulsations through arachnoid granulations in areas of weakened dura overlying congenital bony dehiscences or irregularities in the petroclival fissure. With arachnoid cyst formation and continued, chronic pulsation further bone erosion of the petrous apex occurs creating a smooth, scalloped defect (143,148,151). A similar but distinct entity is a CSF cephalocele of the petrous apex which is a diverticulum of all of the layers of the meninges creating a similar bony defect (173). Isaacson and colleagues hypothesized that a CSF cephalocele may result from increased intracranial pressure transmitting into Meckel’s cave via a patent porus trigeminus (143). In their series three of four patients with a CSF cephalocele were found to have an empty sella, a finding that has been associated with increase intracranial pressure by others (174–176). Three mechanisms have been proposed to explain the expansion of arachnoid cysts. One theory is that intracystic hemorrhage results in the establishment of an osmotic gradient causing subsequent enlargement of the cyst due to fluid shifts. Another mechanism implicates active secretion of fluid by the internal lining of the cyst as the cause of expansion, a theory supported by the finding of Na+ /K+ -ATPase pumps in the cells lining the cyst (177). Further support for this theory was provided by the finding of ectopic choroid plexus within some arachnoid cysts (178). The most widely accepted theory regarding the expansion of arachnoid cysts describes a ballvalve mechanism of CSF trapping within the cyst driven by intermittent increases in intracranial pressure (146,179).
Pathology Histologically arachnoid cysts are CSF spaces surrounded by a thin membrane, a few cell layers thick. The cyst lining resembles normal arachnoid tissue that has been split at its membrane and encloses the fluid cavity (171). The lining consists of pseudostratified epithelial cells with surface microvilli evident on electron microscopy. Rengachary and Watanabe described four histological findings to differentiate the wall of an arachnoid cyst from normal arachnoid as follows: (1) splitting of the arachnoid membrane at the margin of the cyst, (2) a very thick layer of collagen in the cyst wall, (3) the absence of traversing trabecular processes within the cyst, and (4) the presence of hyperplastic arachnoid cells in the cyst wall, which presumably participate in collagen synthesis (180).
Classification Vaquero (181) proposed a classification system for arachnoid cysts of the posterior fossa on anatomic location as
589
follows: laterocerebellar, supracerebellar, retrocerebellar, clival, and mixed. Laterocerebellar cysts occupy the cerebellopontine angle and petrous apex and are generally smaller in size than those seen in other locations. Differentiation of arachnoid cysts in this area from epidermoids requires MRI scanning with fluid-attentuated inversion recovery (FLAIR) and diffusion-weighted image (DWI) sequences (see chap. 10 for imaging characteristics). Retrocerebellar cysts occupy the midline posterior to the cerebellum and can be quite large with significant compression of the cerebellar parenchyma. Supracerebellar arachnoid cysts originate from the quadrigeminal cistern and extend posteriorly along the tentorium with a tendency to cause hydrocephalus. Clival cysts displace the brainstem posteriorly, can extend laterally to the CPA, and are rare.
Clinical Manifestations Arachnoid cysts are highly variable in the spectrum and severity of symptoms that they cause, similar to epidermoids and dermoids. Incidental, asymptomatic arachnoid cysts are encountered commonly since the advent and widespread use of MRI and CT scanning (182,183) in the setting of head trauma and as part of the workup of other neurological complaints. The most frequent presentations of an arachnoid cyst is headache (sometimes associated with nausea and vomiting) along with signs of cerebellar dysfunction including dysmetria and gait disturbance (140). Arachnoid cysts can cause both communicating and noncommunicating hydrocephalus due to compression of the ventricular system or blockage of CSF outflow, causing the typical spectrum of associated symptoms seen in hydrocephalus from other causes. Cysts in the cerebellopontine angle or petrous apex typically present with tinnitus, hearing loss, vertigo, and dizziness (143,146,184–186). Facial paralysis due to an arachnoid cyst has been described but is uncommon (152,187–189). Other unusual presentations include hemifacial spasm (155), trigeminal neuralgia (190), narcolepsy (191), seizures (192), visual loss (143,146), otalgia (146), meningitis, and otorrhea (196). A review by Jallo found that the time of onset of symptoms to diagnosis varied from 4 weeks to 12 months reflecting the vague symptomatology on presentation of these lesions (143).
Radiology Please refer to chapter 4 for a discussion of the imaging characteristics of arachnoid cysts of the skull base. Figures 11– 15 show representative imaging of posterior fossa arachnoid cysts.
Treatment In general, asymptomatic patients with an arachnoid cyst do not require treatment (140,143,194,195) and can be followed clinically and with periodic radiological examinations to monitor for growth or the development of symptoms. However, some authors argue that in the case of a large, asymptomatic lesion the potential for rapid deterioration and even death from intracystic hemorrhage after minor head trauma (196,197) is significant enough to justify operative intervention (198). Indications for surgery to treat arachnoid cysts include demonstrated growth, neural compression, hydrocephalus, or refractory symptoms referable to the lesion (140,144). Surgical treatment options include shunting (cystoperitoneal or ventriculoperitoneal for associated hydrocephalus), fenestration (either open or endoscopic), and microsurgical resection or marsupialization. Cystoperitoneal shunting has the advantage of low
590
Gubbels and Gantz
Figure 13 Axial diffusion-weighted MRI of the same lesion as in Figures 11 and 12. Compare diffusion-weighted imaging characteristics of arachnoid cysts as in this image to those of epidermoids as shown in Figure 4. Figure 11 Axial T1-weighted MRI with gadolinium of a right lateralocerebellar arachnoid cyst.
morbidity and mortality but leads to shunt dependency. Despite this, many authors feel that cystoperitoneal shunting should be the initial surgical procedure for most arachnoid cysts, especially if hydrocephalus is present (134,154,181,199– 202). Ventriculoperitoneal shunting is recommended in cases of communicating hydrocephalus due to an arachnoid cyst (201–203), though not used as the sole treatment modality in the case of posterior fossa lesions due to the risk of upward tentorial herniation with progressive enlargement of the primary lesion (145). Open surgical interventions for arachnoid cysts include fenestration, marsupalization and/or resection. Fenestration has been advocated by many surgeons as it avoids shunt-
Figure 12 Axial constructive interference in steady state (CISS) image of the same arachnoid cyst as in Figure 11.
dependency, allows direct inspection and biopsy of the cyst, and seems to provide long-term results (140,156,204) with few reported failures (154,181,200). However, some authors have reported that cyst fenestration alone does not reliably treat associated hydrocephalus, if present, presumably because there is blockage of the CSF flow in the subarachnoid space potentially due to blood and cellular debris. Because of this many patients who have had cyst fenestration still require ventriculoperitoneal (or lumboperitoneal) shunting to treat continued ventriculomegaly (204,205). In the treatment of CPA epidermoid cysts, some authors advocate resection of the medial, lateral, and posterior walls of the cyst with fenestration of the remaining portion. Advantages to this approach are that this provides optimal treatment by preventing subarachnoid blood and debris from causing future arachnoiditis and obstruction (140,143,203). With this approach, any cyst membrane that may be adherent to brainstem, cerebellum, major
Figure 14 Axial T2-weighted MRI of a retrocerebellar arachnoid cyst.
Chapter 41: Epidermoids, Dermoids, and Other Cysts of the Skull Base
591
REFERENCES
Figure 15 Sagittal T1-weighted MRI with gadolinium of the same lesion as Figure 14.
vessels or cranial nerves is left in place. Arachnoid cysts of the petrous apex can be approached through retrosigmoid, middle fossa, or infracochlear approaches. The infracochlear approach provides decompression of petrous apex cysts only. In contrast, an extradural middle fossa approach has been advocated by many groups as it allows for removal of cyst lining and obliteration of the defect to prevent future infection or CSF leakage in cases of both arachnoid cysts and CSF cephaloceles (143,173,206,207).
Outcome and Prognosis Ciricillo and colleagues reported on a series of 40 pediatric patients with arachnoid cysts over a 10-year period (201). Five patients did not require intervention while 15 underwent cyst fenestration, 12 of whom required shunting for improvement after a median follow-up of 8 years. The remaining 20 patients underwent shunting procedures and 6 of them required subsequent shunt revision. Based on these results, the authors recommended shunting as the initial procedure for children with arachnoid cysts. Levy reported on 39 patients with middle fossa arachnoid cysts that underwent cyst fenestration. Fifteen patients had either hydrocephalus or macrocephaly and required ventriculoperitoneal shunting in addition to fenestration (205). Raffel presented a series of 29 arachnoid cysts in children in the middle and posterior fossae. Fenestration alone was successful in 22 while 7 needed additional cystoperitoneal shunting. Samii and colleagues reported a series of 12 patients with CPA arachnoid cysts treated with cyst resection or maximal fenestration through a suboccipital approach. One patient had seventh and eighth cranial nerve palsies, though all 12 patients were improved symptomatically with this approach with no mortality with 3 years of follow-up (203). Jallo used a similar approach in five pediatric patients with CPA arachnoid cysts including one who had failed previous cystoperitoneal shunting. All five patients were successfully treated using this approach with a mean follow-up of 5 years (140), leading the authors to recommend microsurgical treatment for CPA arachnoid cysts as the optimal initial treatment.
1. Cruveilhier J. Anatomie physiologique du corps humain, vol. II. Paris, France: Balliere, 1829. 2. Bailey P. Cruveilhier’s ‘tumeur perlee. Surg Gynecol Obstet 1920; 31:390–401. 3. Muller J. Uber den feineren Bau und die Formen der krankhaften Geschwulste, vol. 1. Berlin, Germany: G Reimer, 1838. 4. Virchow R. Uber Perlgeschwultse. Virchows Arch A Pathol Anat Histopathol 1855; 8:371–418. 5. Bostroem E. Ueber die pialen epidermoide, dermoide und dermalen dermoide. Zbl Pathol 1897; 8:1–98. 6. Nager GT. Epidermoids involving the temporal bone: Clinical, radiological and pathological aspects. Laryngoscope 1975; 85(12 Pt 2 suppl 2):1–21. 7. Lalwani AK. Meningiomas, epidermoids, and other nonacoustic tumors of the cerebellopontine angle. Otolaryngol Clin North Am 1992; 25(3):707–728. 8. Findeisen L, WT. Uber intracranielle Epidermoide. Zentralbl Neurochir 1937; 2:301–315. 9. Zulch K. Brain tumors: Their biology and pathology, 3rd ed. Berlin, Germany: Springer-Verlag, 1986. 10. Schiffer D. Brain tumors: Biology, pathology and clinical references, 2nd ed. Berlin, Germany: Springer-Verlag, 1997. 11. Berger MS, Wilson CB. Epidermoid cysts of the posterior fossa. J Neurosurg 1985; 62(2):214–219. 12. Netsky MG. Epidermoid tumors. Review of the literature. Surg Neurol 1988; 29(6):477–483. 13. Talacchi A, Sala F, Alessandrini F, et al. Assessment and surgical management of posterior fossa epidermoid tumors: Report of 28 cases. Neurosurgery 1998; 42(2):242–251; discussion 251–252. 14. Kuzeyli K, Duru S, Cakir E, et al. Epidermoid tumor of the occipital bone. Neurosurg Rev 1996; 19(2):109–112. 15. Obrador S, Lopez-Zafra JJ. Clinical features of the epidermoids of the basal cisterns of the brain. J Neurol Neurosurg Psychiatry 1969; 32(5):450–454. 16. Yamakawa K, Shitara N, Genka S, et al. Clinical course and surgical prognosis of 33 cases of intracranial epidermoid tumors. Neurosurgery 1989; 24(4):568–573. 17. Grey PL, Moffat DA, Hardy DG. Surgical results in unusual cerebellopontine angle tumours. Clin Otolaryngol Allied Sci 1996; 21(3):237–243. 18. Bartlett J. Tumors of the lateral ventricle. Except choroid plexus tumors. In: Vinken PJ, Bruyn GW, ed. Handbook of Clinical Neurology, Vol. II. Amsterdam, The Netherlands: North Holland Publishing, 1974:596–609. 19. Lakhdar A, Sami A, Naja A, et al. Epidermoid cyst of the cerebellopontine angle. A surgical series of 10 cases and review of the literature. Neurochirurgie 2003; 49(1):13–24. 20. Chandra PS, Gupta A, Mishra NK, et al. Association of craniovertebral and upper cervical anomalies with dermoid and epidermoid cysts: Report of four cases. Neurosurgery 2005; 56(5):E1155; discussion E1155. 21. El-Bahy K, Kotb A, Galal A, et al. Ruptured intracranial dermoid cysts. Acta Neurochir (Wien) 2006; 148(4):457–462. 22. Abramson RC, Morawetz RB, Schlitt M. Multiple complications from an intracranial epidermoid cyst: Case report and literature review. Neurosurgery 1989; 24(4):574–578. 23. Schijman E, Monges J, Cragnaz R. Congenital dermal sinuses, dermoid and epidermoid cysts of the posterior fossa. Childs Nerv Syst 1986; 2(2):83–89. 24. Alvord EC Jr. Growth rates of epidermoid tumors. Ann Neurol 1977; 2(5):367–370. 25. Boestroem E. Uber die pialen Epidermoide, Dermoide und Lipome und duralen Dermoide. Centralbl allg Path Path Anat 1897; 8:1–98. 26. Scholtz E. Einige Bemerkungen uber das meningaele Cholesteatom in Anschluss an einenFall von Cholestea-tom des III Ventrikels. Virchow’s Arch Path Anat 1906; 184:225– 273. 27. Baumann CH, Bucy PC. Paratrigeminal epidermoid tumors. J Neurosurg 1956; 13(5):455–468.
592
Gubbels and Gantz
28. Fleming JF, Botterell EH. Cranial dermoid and epidermoid tumors. Surg Gynecol Obstet 1959; 109:403–411. 29. Sano K. Intracranial dysembryogenetic tumors: Pathogenesis and their order of malignancy. Neurosurg Rev 2001; 24(4):162– 167; discussion 168–170. 30. Yasargil MG, Abernathey CD, Sarioglu AC. Microneurosurgical treatment of intracranial dermoid and epidermoid tumors. Neurosurgery 1989; 24(4):561–567. 31. Kaylie DM, Warren FM III, Haynes DS, et al. Neurotologic management of intracranial epidermoid tumors. Laryngoscope 2005; 115(6):1082–1086. 32. Kountakis SE, Chang CY, Gormley WB, et al. Migration of intradural epidermoid matrix: Embryologic implications. Otolaryngol Head Neck Surg 2000; 123(3):170–173. 33. Love JG, Kernohan JM. Dermoid and epidermoid tumors (cholesteatomas) of the central nervous system. JAMA 1936; 107:1876–1883. 34. Ulrich J. Intracranial Epidermoids. A study on their distribution and spread. J Neurosurg 1964; 21:1051–1058. 35. Horowitz BL, Chari MV, James R, et al. MR of intracranial epidermoid tumors: Correlation of in vivo imaging with in vitro 13 C spectroscopy. AJNR Am J Neuroradiol 1990; 11(2):299–302. 36. de Souza CE, Sperling NM, da Costa SS, et al. Congenital cholesteatomas of the cerebellopontine angle. Am J Otol 1989; 10(5):358–363. 37. Samii M, Tatagiba M, Piquer J, et al. Surgical treatment of epidermoid cysts of the cerebellopontine angle. J Neurosurg 1996; 84(1):14–19. 38. Vinchon M, Pertuzon B, Lejeune JP, et al. Intradural epidermoid cysts of the cerebellopontine angle: Diagnosis and surgery. Neurosurgery 1995; 36(1):52–56; discussion 56–57. 39. Mohanty A, Venkatrama SK, Rao BR, et al. Experience with cerebellopontine angle epidermoids. Neurosurgery 1997; 40(1):24– 29; discussion 29–30. 40. Sabin HI, Bordi LT, Symon L. Epidermoid cysts and cholesterol granulomas centered on the posterior fossa: Twenty years of diagnosis and management. Neurosurgery 1987; 21(6):798–805. 41. Altschuler EM, Jungreis CA, Sekhar LN, et al. Operative treatment of intracranial epidermoid cysts and cholesterol granulomas: Report of 21 cases. Neurosurgery 1990; 26(4):606–613; discussion 614. 42. De Micheli E, Bricolo A. The long history of a cerebello-pontine angle epidermoid tumour–a case report and lessons learned. Acta Neurochir (Wien) 1996; 138(3):350–354. 43. Antoli-Candela F Jr, Stewart TJ. The pathophysiology of otologic facial paralysis. Otolaryngol Clin North Am 1974; 7(2):309–330. 44. Antonio SM, DK, Del la Cruz A. Epidermoid cysts of the cerebellopontine angle. In: Jackler RK, ed. Neurotology. Philadelphia, PA: Elsevier/Mosby, 2005:841–849. 45. Jamjoom AB, Jamjoom ZA, al-Fehaily M, et al. Trigeminal neuralgia related to cerebellopontine angle tumors. Neurosurg Rev 1996; 19(4):237–241. 46. Eisenman D, VE, Selesnick SH. Unusual tumors of the internal auditory canal and cerebellopontine angle. In: Jackler RK, Driscoll C, eds. Tumors of the Ear and Temporal bone. Philadelphia, PA: Lippincott, Williams & Wilkins, 2000:236–275. 47. Parikh S, Milosevic M, Wong CS, et al. Recurrent intracranial epidermoid cyst treated with radiotherapy. J Neurooncol 1995; 24(3):293–297. 48. Lunardi P, Fortuna A, Cantore G, et al. Long-term evaluation of asymptomatic patients operated on for intracranial epidermoid cysts. Comparison of the diagnostic value of magnetic resonance imaging and computer-assisted cisternography for detection of cholesterin fragments. Acta Neurochir (Wien) 1994; 128(1-4):122–125. 49. Moffat DA, Quaranta N, Baguley DM, et al. Staging and management of primary cerebellopontine cholesteatoma. J Laryngol Otol 2002; 116(5):340–345. 50. Conley F. Epidermoid and dermoid tumors: Clinical features and surgical management. In: Wilkins RH, ed. Neurosurgery. New York: McGraw-Hill, 1985:668–673. 51. Kveton JF, Glasscock ME III, Christiansen SG. Malignant de-
52.
53.
54. 55.
56.
57.
58. 59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72. 73.
74. 75.
generation of an epidermoid of the temporal bone. Otolaryngol Head Neck Surg 1986; 94(5):633–636. Nishio S, Takeshita I, Morioka T, et al. Primary intracranial squamous cell carcinomas: Report of two cases. Neurosurgery 1995; 37(2):329–332. Dubois PJ, Sage M, Luther JS, et al. Case report. Malignant change in an intracranial epidermoid cyst. J Comput Assist Tomogr 1981; 5(3):433–435. Gi H, Yoshizumi H, Nagao S, et al. C-P angle epidermoid carcinoma: A case report. No Shinkei Geka 1990; 18(11):1041–1045. Giangaspero F, Manetto V, Ferracini R, et al. Squamous cell carcinoma of the brain with sarcoma-like stroma. Virchows Arch A Pathol Anat Histopathol 1984; 402(4):459–464. Lewis AJ, Cooper PW, Kassel EE, et al. Squamous cell carcinoma arising in a suprasellar epidermoid cyst. Case report. J Neurosurg 1983; 59(3):538–541. Mori Y, Suzuki Y, Tanasawa T, et al. A case report of epidermoid carcinoma in the cerebello-pontine angle. No Shinkei Geka 1995; 23(10):905–909. Bondeson L, Falt K. Primary intracranial epidermoid carcinoma. Acta Cytol 1984; 28(4):487–489. Landers JW, Danielski JJ. Malignant intracranial epidermoid cyst. Report of a case with leptomeningeal spread. Arch Pathol 1960; 70:419–423. Yamanaka A, Hinohara S, Hashimoto T. Primary diffuse carcinomatosis of the spinal meninges accompanied with a cancerous epidermal cyst of the base of the brain; report of a case of autopsy. Gan 1955; 46(2-3):274–276. Asahi T, Kurimoto M, Endo S, et al. Malignant transformation of cerebello-pontine angle epidermoid. J Clin Neurosci 2001; 8(6):572–574. Ogata N, Jochum W, Aguzzi A, et al. Total removal of a primary intracranial squamous cell carcinoma invading the brain stem. Surg Neurol 1996; 46(5):477–480. Fox H, South EA. Squamous cell carcinoma developing in an intracranial epidermoid cyst (Cholesteatoma). J Neurol Neurosurg Psychiatry 1965; 28:276–281. Knorr JR, Ragland RL, Smith TW, et al. Squamous carcinoma arising in a cerebellopontine angle epidermoid: CT and MR findings. AJNR Am J Neuroradiol 1991; 12(6):1182–1184. Matsuno A, Shibui S, Ochiai C, et al. Primary intracranial epidermoid carcinoma accompanied with epidermoid cyst in the cerebellopontine angle–a case report. No Shinkei Geka 1987; 15(8):851–858. Nishiura I, Koyama T, Handa J, et al. Primary intracranial epidermoid carcinoma–case report. Neurol Med Chir (Tokyo) 1989; 29(7):600–605. Toglia JU, Netsky MG, Alexander E Jr. Epithelial (epidermoid) tumors of the cranium. Their common nature and pathogenesis. J Neurosurg 1965; 23(4):384–393. Lo WWM S-BL. Tumors of the temporal bone and cerebellopontine angle. In: Som P, CH, eds. Head and Neck Imaging. St. Louis: Mosby, 1996:1449–1534. Murase S, Yamakawa H, Ohkuma A, et al. Primary intracranial squamous cell carcinoma–case report. Neurol Med Chir (Tokyo) 1999; 39(1):49–54. Garcia CA, McGarry PA, Rodriguez F. Primary intracranial squamous cell carcinoma of the right cerebellopontine angle. J Neurosurg 1981; 54(6):824–828. Goldman SA, Gandy SE. Squamous cell carcinoma as a late complication of intracerebroventricular epidermoid cyst. Case report. J Neurosurg 1987; 66(4):618–620. Salazar J, Vaquero J, Saucedo G, et al. Posterior fossa epidermoid cysts. Acta Neurochir (Wien) 1987; 85(1–2):34–39. Tamura K, Aoyagi M, Wakimoto H, et al. Malignant transformation eight years after removal of a benign epidermoid cyst: A case report. J Neurooncol 2006; 79(1):67–72. Smith AS. Myth of the mesoderm. AJNR Am J Neuroradiol 1989; 10(2):449. Smith AS, Benson JE, Blaser SI, et al. Diagnosis of ruptured intracranial dermoid cyst: Value MR over CT. AJNR Am J Neuroradiol 1991; 12(1):175–180.
Chapter 41: Epidermoids, Dermoids, and Other Cysts of the Skull Base 76. Stendel R, Pietila TA, Lehmann K, et al. Ruptured intracranial dermoid cysts. Surg Neurol 2002; 57(6):391–398; discussion 398. 77. Verratus. De Bononiensi scientiarum et atrium instituto atque academia comentarii. Bononiae 1745; 2(1):1–184. 78. Lannelongue eA. Traite des Kystes congenitaux. Paris, 1886. 79. Toynbee J. Hairs in the mastoid cells. Trans pathol Soc London 1866; 17:274. 80. Hinton J. Sebaceous tumors and hairs in the tympanum of a boy. Trans pathol Soc London 1866; 17:275. 81. Horrax G. A consideration of the dermal versus the epidermal cholesteatomas having their attachment in the cerebral envelopes. Arch Neur Psychiat 1922; 8:265–285. 82. Pennybacker J, Tytus JS. Pearly tumours in relation to the central nervous system. J Neurol Neurosurg Psychiatry 1956; 19(4):241– 259. 83. Leutner C, Keller E, Pauleit D, et al. An epidermoid of the sphenoid bone and a ruptured intracranial dermoid–a case report. Rofo 1998; 168(2):202–204. 84. Rubin G, Scienza R, Pasqualin A, et al. Craniocerebral epidermoids and dermoids. A review of 44 cases. Acta Neurochir (Wien) 1989; 97(1–2):1–16. 85. Arseni C, Danaila L, Constantinescu AI, et al. Cerebral dermoid tumours. Neurochirurgia (Stuttg) 1976; 19(3):104–114. 86. Guidetti B, Gagliardi FM. Epidermoid and dermoid cysts. Clinical evaluation and late surgical results. J Neurosurg 1977; 47(1):12–18. 87. Maccarty CS, Leavens ME, Love JG, et al. Dermoid and epidermoid tumors in the central nervous system of adults. Surg Gynecol Obstet 1959; 108(2):191–198. 88. Manlapaz JS. Ruptured intracranial dermoid; report of a case and survey of previously reported cases. Am J Surg 1960; 100:723–730. 89. Martinez-Lage JF, Ramos J, Puche A, et al. Extradural dermoid tumours of the posterior fossa. Arch Dis Child 1997; 77(5):427– 430. 90. Akdemir G, Daglioglu E, Ergungor MF. Dermoid lesion of the cavernous sinus: Case report and review of the literature. Neurosurg Rev 2004; 27(4):294–298. 91. Hamer J. Diagnosis by computerized tomography of intradural dermoid with spontaneous rupture of the cyst. Acta Neurochir (Wien) 1980; 51(3-4):219–226. 92. Caldarelli M, Colosimo C, Di Rocco C. Intra-axial dermoid/epidermoid tumors of the brainstem in children. Surg Neurol 2001; 56(2):97–105. 93. Fried MP, Vernick DM. Dermoid cyst of the middle ear and mastoid. Otolaryngol Head Neck Surg 1984; 92(5):594– 596. 94. Vrabec JT, Schwaber MK. Dermoid tumor of the middle ear: Case report and literature review. Am J Otol 1992; 13(6):580– 581. 95. Behnke EE, Schindler RA. Dermoid of the petrous apex. Laryngoscope 1984; 94(6):779–783. 96. Lunardi P, Missori P, Gagliardi FM, et al. Dermoid cysts of the posterior cranial fossa in children. Report of nine cases and review of the literature. Surg Neurol 1990; 34(1):39–42. 97. Roberts AP. A case of intracranial dermoid cyst associated with the Klippel-Feil deformity and recurrent meningitis. Arch Dis Child 1958; 33(169):222–225. 98. Whittle IR, Besser M. Congenital neural abnormalities presenting with mirror movements in a patient with Klippel-Feil syndrome. Case report. J Neurosurg 1983; 59(5):891–894. 99. Diekmann-Guiroy B, Huang PS. Klippel-Feil syndrome in association with a craniocervical dermoid cyst presenting as aseptic meningitis in an adult: Case report. Neurosurgery 1989; 25(4):652–655. 100. Dickey W, Hawkins SA, Kirkpatrick DH, et al. Posterior fossa dermoid cysts and the Klippel-Feil syndrome. J Neurol Neurosurg Psychiatry 1991; 54(11):1016–1017. 101. Kuribayashi K, Nakasu S, Matsumura K, et al. Dermoid cyst in the fourth ventricle associated with Klippel-Feil syndrome. No To Shinkei 1993; 45(8):747–751. 102. Kennedy PT, McAuley DJ. Association of posterior fossa der-
103.
104.
105.
106.
107. 108.
109.
110. 111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122. 123.
124.
125.
593
moid cyst and Klippel-Feil syndrome. AJNR Am J Neuroradiol 1998; 19(1):195; discussion 196. Hinojosa M, Tatagiba M, Harada K, et al. Dermoid cyst in the posterior fossa accompanied by Klippel-Feil syndrome. Childs Nerv Syst 2001; 17(1-2):97–100. Sharma MS, Sharma BS, Yadav A, et al. Posterior fossa dermoid in association with Klippel-Feil syndrome–a short report. Neurol India 2001; 49(2):210–212. Murphy MJ, Risk WS, VanGilder JC. Intracranial dermoid cyst in Goldenhar’s syndrome. Case report. J Neurosurg 1980; 53(3):408–410. Shirakuni T, Yamasaki S, Sato H. An infant case of Goldenhar syndrome associated with intracranial lesions. No To Hattatsu 1985; 17(3):251–255. Wilson GN. Cranial defects in the Goldenhar syndrome. Am J Med Genet 1983; 14(3):435–443. Lunardi P, Missori P, Rizzo A, et al. Chemical meningitis in ruptured intracranial dermoid. Case report and review of the literature. Surg Neurol 1989; 32(6):449–452. Hyams V, Batsakis J, Michaels L. Tumors of the Upper Respiratory Tract and Ear, vol. 25. Washington D. C.: Armed Forces Institute of Pathology, 1988:201–206. Howie TO. A case of dermoid or developmental cyst of the middle-ear cavity. J Laryngol Otol 1962; 76:62–66. Logue V, Till K. Posterior fossa dermoid cysts with special reference to intracranial infection. J Neurol Neurosurg Psychiatry 1952; 15(1):1–12. Nakagawa K, Ohno K, Nojiri T, et al. Interdural dermoid cyst of the cavernous sinus presenting with oculomotor palsy: Case report. No Shinkei Geka 1997; 25(9):847–851. North KN, Antony JH, Johnston IH. Dermoid of cavernous sinus resulting in isolated oculomotor nerve palsy. Pediatr Neurol 1993; 9(3):221–223. Phillips WE II, Martinez CR, Cahill DW. Ruptured intracranial dermoid tumor secondary to closed head trauma. Computed tomography and magnetic resonance imaging. J Neuroimaging 1994; 4(3):169–170. Carvalho GA, Cervio A, Matthies C, et al. Subarachnoid fat dissemination after resection of a cerebellopontine angle dysontogenic cyst: Case report and review of the literature. Neurosurgery 2000; 47(3):760–763; discussion 763–764. Cavazzani P, Ruelle A, Michelozzi G, et al. Spinal dermoid cysts originating intracranial fat drops causing obstructive hydrocephalus: Case reports. Surg Neurol 1995; 43(5):466–469; discussion 469–470. Karabulut N, Oguzkurt L. Tetraventricular hydrocephalus due to ruptured intracranial dermoid cyst. Eur Radiol 2000; 10(11):1810–1811. Larsson EM, Brandt L, Holtas S. Persisting intraventricular fatfluid levels following surgery on a ruptured dermoid cyst of the posterior fossa. Acta Radiol 1987; 28(4):489–490. Martin R, Knone A, Schuknecht B, et al. Rapid development of occlusion hydrocephalus by intraventricular fat possibly derived from a ruptured dermoid cyst. J Neurol Neurosurg Psychiatry 1989; 52(1):134–135. Ahmad I, Tominaga T, Ogawa A, et al. Ruptured suprasellar dermoid associated with middle cerebral artery aneurysm: Case report. Surg Neurol 1992; 38(5):341–346. Ford K, Drayer B, Osborne D, et al. Case report. Transient cerebral ischemia as a manifestation of ruptured intracranial dermoid cyst. J Comput Assist Tomogr 1981; 5(6):895– 897. Mikhael MA. Transient spasm of carotid siphon complicating ruptured cranial dermoid cyst. Radiology 1982; 144(4):824. Ecker RD, Atkinson JL, Nichols DA. Delayed ischemic deficit after resection of a large intracranial dermoid: Case report and review of the literature. Neurosurgery 2003; 52(3):706–710; discussion 709–710. Messori A, Polonara G, Serio A, et al. Expanding experience with spontaneous dermoid rupture in the MRI era: Diagnosis and follow-up. Eur J Radiol 2002; 43(1):19–27. Oursin C, Wetzel SG, Lyrer P, et al. Ruptured intracranial der-
594
126.
127.
128.
129. 130. 131.
132. 133. 134. 135. 136.
137. 138. 139.
140.
141.
142.
143.
144.
145. 146. 147.
148.
149. 150. 151.
152.
Gubbels and Gantz moid cyst. J Neurosurg Sci 1999; 43(3):217–220; discussion 220– 221. Lepintre J, Labrune M. Congenital dermal fistulae communicating with the central nervous system. 21 operated cases in children. Neurochirurgie 1970; 16(4):335–348. Hamel E, Frowein RA, Karimi-Nejad A. Intracranial intradural epidermoids and dermoids. Surgical results of 38 cases. Neurosurg Rev 1980; 3(4):215–219. Bright R. Diseases of the Brain and Nervous System, Part I. London: Longman, Resse, Orme, Brown and Green, 1831:437– 439. Maunsell H. Subtentorial hydatid tumor removed by trephining: Recovery. N Z Med J 1889; 2:151–156. Mullin W. Circumscribed arachnoid cyst giving symptoms of an acoustic neuroma. Trans Am Otol Soc 1932; 22:72–101. Aubry M, Ombredanne M. Etude Oto-Neurologique et Chirurgicale du Vertige. Indications et Resultats De la Chirurgie Intracranienne Du Nerf Auditif. Paris: Masson and Cie, 1937:100– 122. Lehman RA, Fieger HG Jr. Arachnoid cyst producing recurrent neurological disturbances. Surg Neurol 1978; 10(2):134–136. Alker GJ Jr, Glasauer FE, Leslie EV. Radiology of a large cisterna magna cyst. A case report. Arch Neurol 1979; 36(6):376–379. di Rocco C, Caldarelli M, di Trapani G. Infratentorial arachnoid cysts in children. Childs Brain 1981; 8(2):119–133. Handa J, Nakano Y, Aii H. CT cisternography with intracranial arachnoidal cysts. Surg Neurol 1977; 8(6):451–454. McDonald JV, Colgan J. Arachnoidal cysts of posterior fossa. Their pneumoencephalographic appearance: A case report. Neurology 1964; 14:643–646. Williams B. Subarachnoid pouches of the posterior fossa with syringomyelia. Acta Neurochir (Wien) 1979; 47(3-4):187–217. Pappas DG, Brackmann DE. Arachnoid cysts of the posterior fossa. Otolaryngol Head Neck Surg 1981; 89(2):328–332. Gomez MR, Yanagihara T, MacCarty CS. Arachnoid cyst of the cerebellopontine angle and infantile spastic hemiplegia. Case report. J Neurosurg 1968; 29(1):87–90. Jallo GI, Woo HH, Meshki C, et al. Arachnoid cysts of the cerebellopontine angle: Diagnosis and surgery. Neurosurgery 1997; 40(1):31–37; discussion 37–38. Buongiorno G, Ricca G. Supratentorial arachnoid cyst mimicking a Meniere’s disease attack. J Laryngol Otol 2003; 117(9):728– 730. Thijssen HO, Marres EH, Slooff JL. Arachnoid cyst simulating intrameatal acoustic neuroma. Neuroradiology 1976; 11(4):205– 207. Isaacson B, Coker NJ, Vrabec JT, et al. Invasive cerebrospinal fluid cysts and cephaloceles of the petrous apex. Otol Neurotol 2006; 27(8):1131–1141. Kandenwein JA, Richter HP, Borm W. Surgical therapy of symptomatic arachnoid cysts – an outcome analysis. Acta Neurochir (Wien) 2004; 146(12):1317–1322; discussion 1322. Sinha S, Brown JI. Familial posterior fossa arachnoid cyst. Childs Nerv Syst 2004; 20(2):100–103. Bohrer PS, Chole RA. Unusual lesions of the internal auditory canal. Am J Otol 1996; 17(1):143–149. Cheung SW, Broberg TG, Jackler RK. Petrous apex arachnoid cyst: Radiographic confusion with primary cholesteatoma. Am J Otol 1995; 16(5):690–694. Batra A, Tripathi RP, Singh AK, et al. Petrous apex arachnoid cyst extending into Meckel’s cave. Australas Radiol 2002; 46(3):295– 298. Schuknecht HF, Gao YZ. Arachnoid cyst in the internal auditory canal. Ann Otol Rhinol Laryngol 1983; 92(6Pt 1):535–541. Abbott R. The endoscopic management of arachnoidal cysts. Neurosurg Clin N Am 2004; 15(1):9–17. Falcioni M, Caruso A, Taibah A, et al. Arachnoid cysts of the petrous apex in a patient with vestibular schwannoma. Otolaryngol Head Neck Surg 2000; 123(5):657–658. Kacker A, Bent JP, Abbott R. Non-communicating arachnoid cyst of the temporal bone presenting as facial nerve paralysis in an infant. Int J Pediatr Otorhinolaryngol 1999; 49(2):151–154.
153. Wester K. Peculiarities of intracranial arachnoid cysts: Location, sidedness, and sex distribution in 126 consecutive patients. Neurosurgery 1999; 45(4):775–779. 154. Galassi E, Tognetti F, Frank F, et al. Infratentorial arachnoid cysts. J Neurosurg 1985; 63(2):210–217. 155. Higashi S, Yamashita J, Yamamoto Y, et al. Hemifacial spasm associated with a cerebellopontine angle arachnoid cyst in a young adult. Surg Neurol 1992; 37(4):289–292. 156. Hoffman HJ, Hendrick EB, Humphreys RP, et al. Investigation and management of suprasellar arachnoid cysts. J Neurosurg 1982; 57(5):597–602. 157. Allen A, Wiegmann TB, MacDougall ML. Arachnoid cyst in a patient with autosomal-dominant polycystic kidney disease. Am J Kidney Dis 1986; 8(2):128–130. 158. Kara B, Kayserili H, Imer M, et al. Quadrigeminal cistern arachnoid cyst in a patient with Kabuki syndrome. Pediatr Neurol 2006; 34(6):478–480. 159. Hajje MJ, Nachanakian A, Haddad J. Goldenhar syndrome and arachnoid cyst. Arch Pediatr 2003; 10(4):353–354. 160. Welch KO, Tekin M, Nance WE, et al. Chudley-McCullough syndrome: Expanded phenotype and review of the literature. Am J Med Genet A 2003; 119(1):71–76. 161. Masuno M, Fukushima Y, Sugio Y, et al. Partial distal 12q trisomy with arachnoid cyst. Jinrui Idengaku Zasshi 1987; 32(1):39–43. 162. Balci S, Oguz KK. Cri-du-chat syndrome associated with arachnoid cyst causing triventricular hydrocephalus. Clin Dysmorphol 2001; 10(4):289–290. 163. Hald JK, Nakstad PH, Skjeldal OH, et al. Bilateral arachnoid cysts of the temporal fossa in four children with glutaric aciduria type I. AJNR Am J Neuroradiol 1991; 12(3):407–409. 164. Martinez-Lage JF, Casas C, Fernandez MA, et al. Macrocephaly, dystonia, and bilateral temporal arachnoid cysts: Glutaric aciduria type 1. Childs Nerv Syst 1994; 10(3):198–203. 165. Jamjoom ZA, Okamoto E, Jamjoom AH, et al. Bilateral arachnoid cysts of the sylvian region in female siblings with glutaric aciduria type I. Report of two cases. J Neurosurg 1995; 82(6):1078–1081. 166. Renner C, Razeghi S, Uberall MA, et al. Clinically asymptomatic glutaric aciduria type I in a 4 5/12-year-old girl with bilateral temporal arachnoid cysts. J Inherit Metab Dis 1997; 20(6):840– 841. 167. Lutcherath V, Waaler PE, Jellum E, et al. Children with bilateral temporal arachnoid cysts may have glutaric aciduria type 1 (GAT1); operation without knowing that may be harmful. Acta Neurochir (Wien) 2000; 142(9):1025–1030. 168. Handa J, Okamoto K, Sato M. Arachnoid cyst of the middle cranial fossa: Report of bilateral cysts in siblings. Surg Neurol 1981; 16(2):127–130. 169. Pomeranz S, Constantini S, Lubetzki-Korn I, et al. Familial intracranial arachnoid cysts. Childs Nerv Syst 1991; 7(2):100–102. 170. Tolmie JL, Day R, Fredericks B, et al. Dominantly inherited cerebral dysplasia: Arachnoid cyst associated with mild mental handicap in a mother and her son. J Med Genet 1997; 34(12):1018–1020. 171. Nager G. Epidermoids (Congenital Cholesteatomas). In: Nager G, ed. Pathology of the Ear and Temporal Bone, vol. I. Baltimore, MD: Williams & Wilkins, 1993:710–742. 172. Starkman SP, Brown TC, Linell EA. Cerebral arachnoid cysts. J Neuropathol Exp Neurol 1958; 17(3):484–500. 173. Moore KR, Fischbein NJ, Harnsberger HR, et al. Petrous apex cephaloceles. AJNR Am J Neuroradiol 2001; 22(10):1867– 1871. 174. Prichard CN, Isaacson B, Oghalai JS, et al. Adult spontaneous CSF otorrhea: Correlation with radiographic empty sella. Otolaryngol Head Neck Surg 2006; 134(5):767–771. 175. Foley KM, Posner JB. Does pseudotumor cerebri cause the empty sella syndrome? Neurology 1975; 25(6):565–569. 176. Weisberg LA, Housepian EM, Saur DP. Empty sella syndrome as complication of benign intracranial hypertension. J Neurosurg 1975; 43(2):177–180. 177. Go KG, Houthoff HJ, Blaauw EH, et al. Arachnoid cysts of the
Chapter 41: Epidermoids, Dermoids, and Other Cysts of the Skull Base
178.
179.
180.
181. 182. 183.
184.
185.
186.
187.
188.
189. 190.
191.
192.
sylvian fissure. Evidence of fluid secretion. J Neurosurg 1984; 60(4):803–813. Schuhmann MU, Tatagiba M, Hader C, et al. Ectopic choroid plexus within a juvenile arachnoid cyst of the cerebellopontine angle: Cause of cyst formation or reason of cyst growth. Pediatr Neurosurg 2000; 32(2):73–76. Tirakotai W, Schulte DM, Bauer BL, et al. Neuroendoscopic surgery of intracranial cysts in adults. Childs Nerv Syst 2004; 20(11-12):842–851. Rengachary SS, Watanabe I. Ultrastructure and pathogenesis of intracranial arachnoid cysts. J Neuropathol Exp Neurol 1981; 40(1):61–83. Vaquero J, Carrillo R, Cabezudo JM, et al. Arachnoid cysts of the posterior fossa. Surg Neurol 1981; 16(2):117–121. Garcia-Bach M, Isamat F, Vila F. Intracranial arachnoid cysts in adults. Acta Neurochir Suppl (Wien) 1988; 42:205–209. Eskandary H, Sabba M, Khajehpour F, et al. Incidental findings in brain computed tomography scans of 3000 head trauma patients. Surg Neurol 2005; 63(6):550–553; discussion 553. Hadley MN, Grahm TW, Daspit CP, et al. Otolaryngologic manifestations of posterior fossa arachnoid cysts. Laryngoscope 1985; 95(6):678–681. Haberkamp TJ, Monsell EM, House WF, et al. Diagnosis and treatment of arachnoid cysts of the posterior fossa. Otolaryngol Head Neck Surg 1990; 103(4):610–614. Pollice PA, Bhatti NI, Niparko JK. Imaging quiz case 1. Posterior fossa arachnoid cyst. Arch Otolaryngol Head Neck Surg 1997; 123(7):762, 764–765. Sumner TE, Benton C, Marshak G. Arachnoid cyst of the internal auditory canal producing facial paralysis in a three-year-old child. Radiology 1975; 114(2):425–426. Bonora G, Manzoni D, Frattini D, et al. Facial paralysis secondary to arachnoid cysts. Description of a case in a girl with precocious puberty. Pediatr Med Chir 1989; 11(3):343–345. Diwan AG, Agarwal SI, Lakade SM, et al. Facial paralysis due to arachnoid cyst. J Assoc Physicians India 2005; 53:544. Babu R, Murali R. Arachnoid cyst of the cerebellopontine angle manifesting as contralateral trigeminal neuralgia: Case report. Neurosurgery 1991; 28(6):886–887. Nakano H, Ogashiwa M. Complete remission of narcolepsy after surgical treatment of an arachnoid cyst in the cerebellopontine angle. J Neurol Neurosurg Psychiatry 1995; 58(2): 264. Gelabert-Gonzalez M. Intracranial arachnoid cysts. Rev Neurol 2004; 39(12):1161–1166.
595
193. Abramson M, McMurtry JG, Kadis G. Combined otoneurosurgical approach to patients with chronic ear disease and cerebrospinal otorrhea. Laryngoscope 1979; 89(6Pt 1):918–921. 194. Becker T, Wagner M, Hofmann E, et al. Do arachnoid cysts grow? A retrospective CT volumetric study. Neuroradiology 1991; 33(4):341–345. 195. Punzo A, Conforti R, Martiniello D, et al. Surgical indications for intracranial arachnoid cysts. Neurochirurgia (Stuttg) 1992; 35(2):35–42. 196. Ehrensberger J, Gysler R, Illi OE, et al. Congenital intracranial cysts: Clinical findings, diagnosis, treatment and follow-up. A multicenter, retrospective long-term evaluation of 72 children. Eur J Pediatr Surg 1993; 3(6):323–334. 197. McCullough DC, Harbert JC, Manz HJ. Large arachnoid cysts at the cranial base. Neurosurgery 1980; 6(1):76–81. 198. Chao TK. Middle cranial fossa arachnoid cysts causing sensorineural hearing loss. Eur Arch Otorhinolaryngol 2005; 262(11):925–927. 199. Mason TB II, Chiriboga CA, Feldstein NA, et al. Massive intracranial arachnoid cyst in a developmentally normal infant: Case report and literature review. Pediatr Neurol 1997; 16(1):59– 62. 200. Anderson FM, Segall HD, Caton WL. Use of computerized tomography scanning in supratentorial arachnoid cysts. A report on 20 children and four adults. J Neurosurg 1979; 50(3):333– 338. 201. Ciricillo SF, Cogen PH, Harsh GR, et al. Intracranial arachnoid cysts in children. A comparison of the effects of fenestration and shunting. J Neurosurg 1991; 74(2):230–235. 202. Harsh GRT, Edwards MS, Wilson CB. Intracranial arachnoid cysts in children. J Neurosurg 1986; 64(6):835–842. 203. Samii M, Carvalho GA, Schuhmann MU, et al. Arachnoid cysts of the posterior fossa. Surg Neurol 1999; 51(4):376–382. 204. Raffel C, McComb JG. To shunt or to fenestrate: Which is the best surgical treatment for arachnoid cysts in pediatric patients? Neurosurgery 1988; 23(3):338–342. 205. Levy ML, Meltzer HS, Hughes S, et al. Hydrocephalus in children with middle fossa arachnoid cysts. J Neurosurg 2004; 101 (1 Suppl):25–31. 206. Schick B, Draf W, Kahle G, et al. Occult malformations of the skull base. Arch Otolaryngol Head Neck Surg 1997; 123(1):77– 80. 207. Mulcahy MM, McMenomey SO, Talbot JM, et al. Congenital encephalocele of the medial skull base. Laryngoscope 1997; 107(7):910–914.
42 Fibro-Osseous Lesions of the Skull Base Ian T. Jackson
tion at the woven bone stage; why this occurs remains a mystery. Another feature is that the growth can be sporadic and once again the reason for this is unknown but changes in the level of growth hormone as a progenitor have been postulated. Another characteristic is that there is not much growth after the age of 30 in most cases but it should be noted that there is frequently a growth spurt in the teenage years. However, in our series growth has occurred at many different times and it has shown assorted types of growth. In our practice, which is mainly restricted to the craniofacial area, temporo-orbital involvement with its associated problems is seen most frequently (4). This situation consists of proptosis, vertical globe displacement, optic nerve compression, cranial nerve compression, and narrowing of the superior orbital fissure (Fig. 1). These changes can have significant repercussions. The most serious one is blindness; the other orbital problems are diplopia, nerve compression in the superior orbital fissure, and epiphora due to lacrimal duct compression. With maxillary involvement, the cheek is displaced, the globe may be displaced in any dimension, again resulting in diplopia, and once again, epiphora is a problem. Another feature due to a decrease in orbital volume caused by increasing involvement of the walls is proptosis. This is frequently associated with increasing size of the frontal bone together with the orbital roof, and this may cause the eye to be displaced inferiorly. When blindness occurs, it is caused by a spontaneous hemorrhage into the cystic maxillary and/or orbital bones; this cystic change is again a result of the fibrous dysplasia. Another cause of blindness is compression of the optic nerve due to involvement of the optic nerve canal causing narrowing. This occurs suddenly without any advance warning, it is not a gradual deterioration. Because of this it is of vital importance that the optic nerve canal is inspected carefully whenever a CT scan is taken. It is our strong opinion that a canal, which is decreasing in size, should be released through an intracranial approach: other corrective procedures can be performed at the same time. In the six cases in which we have decompressed the optic nerve canal there have been no problems so far such as postoperative deterioration of vision; also, there has been no improvement in vision. It is of vital importance that the contouring burr is cooled when thinning the bone to ensure that there is not enough heat generated to injure the nerve. The thinned roof of the optic nerve canal is then removed with fine ronguers, as are the lateral walls. On some occasions the roof of the canal can be fractured using a narrow periosteal elevator. At the preoperative assessment session, it should be clearly understood by both surgeon and patient that an improvement in vision will be unlikely to result from the intervention, no matter what is done or what instruments are used. This procedure is purely prophylactic. Having said
Fibro-osseous lesions of the skull base are challenging in every way, mainly due to their anatomical situation. Exposure can be difficult as can the resection. Without care, knowledge, and experience, disasters can occur. In many cases, accurate diagnosis, good preoperative planning, and close cooperation with an experienced neurosurgeon are essential. It is important that the surgeons know exactly what part each plays in the surgical procedure. If this kind of arrangement is not in place, these procedures can become chaotic and the patient suffers. Experienced nurses are an essential part of the team and they must be told this and complimented. They can be very helpful when serious complications occur; fortunately, this is a rarity. Skull base and craniofacial surgery can be stressful and, in some cases, lengthy and thus it is important to have a harmonious and well-organized operating room environment. The approaches to the skull base have evolved from the classic rules and approaches to head and neck tumors with the addition of experience gained from the correction of major congenital craniofacial deformities. As a result of this, some new approaches have been developed and applied to resect skull base tumors. This has allowed us to approach these difficult problems, with more confidence of achieving complete resection followed most times by a satisfactory reconstruction. The philosophy is one of disassembly for exposure, followed by resection, reconstruction, and re-assembly.
FIBROUS DYSPLASIA OF THE CRANIOFACIAL REGION This benign fibro-osseous condition can arise in any part of the facial skeleton (1,2). It usually occurs at an early age, the classical history being one of a slowly enlarging area of bone. The cause of the condition is unknown. It may be totally asymptomatic or it can interfere with function. It can also cause significant craniofacial deformity. Very occasionally there may be associated pain, and sometimes tenderness. The basic skeletal problem is that mature bone is replaced by immature bone and fibrous tissue. This usually occurs at a single site, the most frequent being the cranial base, the maxilla, or the mandible. The polyostotic variant is not common and this usually occurs in patients who have the McCune–Albright syndrome (3). In this latter condition, there are endocrine problems, long bone deformities, skin pigmentation, and fibrous dysplasia isolated to the facial skeleton. The diagnosis is made on examination of the x-rays and CT scans. The histological appearance is that of a variable amount of fibrous dysplasia with some of the intervening bone having a characteristic shape, this is termed Chinese characters. This situation has all the features of an arrest of matura597
598
Jackson
(A)
(C)
(B)
(D)
Figure 1 (A, B) Involvement of the orbital wall with fibrous dysplasia. The right orbit is enlarged circumferentially causing proptosis and blindness. There is also significant invasion of the temporal bone. (C, D) The three-dimensional reconstruction shows the orbital deformity and the full-thickness defect in the posterior orbit.
this, however, there has been the occasional report of the sight improving in patients treated by steroid medications and surgery after having had a sudden loss of vision; however, as stated previously, we have not ourselves experienced this (4,5). If there is a rapid loss of vision, this should be considered as an emergency situation. Large doses of steroids should be given followed by immediate optic nerve decompression as indicated above. The other problem of craniofacial fibrous dysplasia is the deformity, which results from it. There is always the question of how best to treat this situation. It is probably advis-
able, in the majority of asymptomatic cases, to simply contour the involved areas (6–8). This is carried out by exposing the upper face and orbit through a coronal flap. The involved maxilla and/or the mandible are approached through an intraoral approach or midface approach (9). When there is extensive involvement with gross anatomical disruption, particularly with cystic change, total resection should be performed (Fig. 2). In almost every area, the subsequent reconstruction can be performed using split skull grafts. Any residual fullthickness skull defects will be reconstructed with bone dust and bone fragments stabilized with Surgicel moistened with
Chapter 42: Fibro-Osseous Lesions of the Skull Base
(A)
599
(B)
(D) (C)
Figure 2 (A) There is replacement of the whole right maxilla with fibrous dysplasia. (B, C) Exposure is obtained by a Weber–Fergusson approach and total resection of the involved maxilla is performed. (D) The defect was reconstructed using split cranial bone grafts.
saline. In large defects, a portion of full thickness skull equal in size of that of the defect is harvested and split. One portion reconstructs the resection site and the other is placed back into the donor area. These grafts are stabilized with the appropriate plates and screws. The approach for fronto-orbital involvement is by a coronal flap, a frontal craniotomy, resection, and reconstruction (Fig. 3). The latter must seal off the nasal cavity
using free tissue transfer or a galeofrontalis flap. Holes are drilled around the defect in the floor of the anterior cranial fossa, and the vascular transfer is sutured in an overlapping position. This ensures that when the frontal lobe resumes their correct position the connection down into the nose is securely sealed off (Fig. 4). Unfortunately, the patient with fibrous dysplasia is never cured, they must be followed up and they may require
600
Jackson
(A)
(B)
(C)
Figure 3 (A) This patient had extensive involvement of the orbits and anterior skull base. (B) A significant resection is performed to decompress the orbit and optic nerve. Following this a galeofrontalis flap seals off the nasal cavity and the orbital reconstruction with bone graft is completed. (C) The end result is an improvement but further surgery may be necessary.
further surgery because of deformity or functional problems. There has been an attempt to at least slow down the progression of the disease with systemic treatment with bisphosphonates (10). It has been our impression that in some cases this may have slowed the progress of the disease and in some cases the pain, if present, has been relieved to some extent. In the majority of cases there has been little response.
In the McCune–Albright syndrome there is significant involvement of the skull, skull base, and upper and lower jaws with fibrous dysplasia (11). The orbit is extensively involved as is the optic nerve canal. There is pituitary hyperplasia and gigantism. In addition to any bony surgery for appearance the optic nerve canal may need to be enlarged to decompress the optic nerve. This is always a difficult decision
Chapter 42: Fibro-Osseous Lesions of the Skull Base
601
(A)
(B)
(C)
(D)
Figure 4 (A) The involvement in this case includes the floor of the anterior cranial fossa, the left orbit, and nasal cavity. (B, C) Through a frontonasal approach, the lesion is totally resected. (D) Reconstruction is performed using an inferiorly based galeofrontalis flap. (E, F) This overlays the central anterior cranial fossa defect and is sutured through drill holes around the floor defect. (G) When the brain comes forward and downward, a secure seal is formed. The end result is satisfactory and stable. (E, F, and G shown on page 596).
602
Jackson
(E)
(F)
(G)
Figure 4 (Continued)
to make. As stated earlier, there is never a gradual decrease in vision, it is all or nothing thus the decision to actively treat the condition is made on the CT scan appearance (12). Through a coronal approach with removal of the superior orbit and a middle cranial fossa exposure, the canal is gradually de-roofed. This is complicated by the variable thickness of the involved bone. If possible, lateral and medial decompression is also undertaken (Fig. 5). The potential problem in these cases is blindness—in all the fibrous dysplasia cases treated only six had optic nerve compression, only two had McCune–Albright syndrome. There was no visual damage, there was no improvement in vision, and there has been no deterioration of vision on long-term follow-up. An attempt was made, in the past, to slow up the growth of the condition by the use of radiation therapy. Unfortunately, this generated a change of the lesion into an osteogenic sarcoma (13–15). In our practice, we have had only one such case. This was treated in Eastern Europe, and the patient presented with a significant tumor, which was resected and reconstructed. The follow-up, by e-mail, has been very satisfactory with no recurrence over the long term.
NEUROFIBROMATOSIS This is presented because of the significant bony deformities, which are associated with this condition. Neurofibromatosis is an inherited trait; it is autosomally dominant with almost complete penetrance. The condition occurs in 1 in 3000 live births. In neurofibromatosis 1 (NF1) there are defects on chromosome 17. In order to make a diagnosis of neurofibromatosis 1 (NF1), two or more criteria must be fulfilled. These criteria are as follows: (i) Six or more caf´e au lait spots over 5 mm in diameter in prepubertal patients and greater than 15 mm in postpubertal patients; (ii) axillary or groin freckles; (iii) two or more neurofibroma or one plexiform neurofibroma of any type; (iv) an optic glioma or two or more Lisch nodules; (v) a distinctive bone lesion, e.g., sphenoid dysplasia, thinning of long bone cortex, and a first degree relative with NF1; (vi) ophthalmic problems, choroid hematomas, prominent corneal nerves, congenital glaucoma, choroidal caf´e au
lait spots, neurofibroma of orbit or lids, absence of the greater wing of the sphenoid (16). The most common nonmalignant eyelid tumor in children is the plexiform neurofibroma. The incidence is between 4% and 5%, and the sensory nerves are most frequently affected. The characteristic sign is the S-shaped lower eyelid due to lateral lid involvement, which presents as a swelling. When this area is explored, it seems like a bag of worms. Sometimes the external appearance shows a similar irregularity especially in the upper eyelid. This slow-growing lesion is most frequently found in children from 1 to 5 years. Although growth is slow in most cases there may be periods of increased activity. It is rare to have malignant change in these patients, but it has been reported (17). The CT scan reveals a soft tissue mass in the involved area which can be in the eyelids, the orbit, the cheek, or it may extend intracranially. In this condition, fortunately, sarcoma is rare. As pointed out earlier, on CT scan, there is sphenoid wing absence. The main soft tissue mass is frequently in the temporal area and it may extend through the enlarged superior or inferior orbital fissures. The expanding mass of the lesion causes the orbital volume to be increased with enlargement of the bony orbit. On 3-D CT scan the orbit is egg-shaped. Resection can be bloody, and it has been stated that it is often incomplete and recurrence is frequent (18) (Fig. 6). The swelling takes a long time to resolve, and multiple surgeries are required. It will be seen from our experience that this is not the case in the majority of patients, which have been treated by our team, this may result from the radical approach to this condition (19). Medical treatment has not been very helpful. The medications being tried are ketotifen fumarate, a mast cell stabilizer, a combination of Interferon and CS retinoic acid and thalidomide. Resection with the CO2 laser has also been tried without much success. The typical patient referred to our practice has a characteristic involvement of the facial area; this has allowed three categories of severity to be described that, in turn, indicate the treatment required. The lesion is most frequently temporoorbital in distribution. In the temporal area, there is expansion due to the neurofibromatous tissue with noticeable increasing soft tissue and temporal bone protrusion outwards, with time there may be bony erosion with resulting defects.
Chapter 42: Fibro-Osseous Lesions of the Skull Base
(A)
603
(B)
(C)
Figure 5 (A, B) This patient illustrates an example of a severe McCune–Albright syndrome. In addition to the significant involvement of the skull, left orbit, left nose, and left maxilla with fibrous dysplasia, the patient is over 7 feet tall. (C) The main intervention was to decompress his optic nerve as seen in the last figure.
(A)
(B)
Figure 6 (A) This young boy has orbital neurofibromatosis. The eyelids, cheek, and orbit are involved. (B) The characteristic egg-shaped bony orbit with an absent posteromedial orbital wall is well illustrated as is the increase in orbital volume.
604
Jackson
(A)
(C)
(B)
Figure 7 (A) This patient, with neurofibrosarcoma, presented with a recurrence post-resection, post-irradiation. (B) A radial resection using a coronal flap exposure was reconstructed with a temporalis muscle with a galeal extension. (C) This allowed good healing to be achieved.
These defects follow the pattern of the temporal lobe gyri. The orbit is significantly involved; characteristically the sphenoid wing may be absent and/or the fissures, especially the superior, may be considerably increased in size. This is often associated with eye pulsation. The orbit is typically large and egg-shaped but, in spite of this, there may be proptosis. The eye may or may not be functional. The eyelids are involved; as stated earlier, these may be like a bag of worms. They are excessive in all layers due to expansion by the tumor and there is upper eyelid ptosis. The eye may be large but may have vision or it may be blind. The latter occurs in proportion to the eye deformity, this helps greatly in decision-making and when there are discussions
with the patient on whether to proceed to exenteration. The lesion often extends into the lower lid and down into the cheek. The investigations required are an assessment of vision, CT scan, and very frequently a 3-D CT scan. This is used to plan the surgery in terms of resection, temporo-orbital reconstruction, both bony and soft tissue. As mentioned above, the eye frequently has little or no vision (Fig. 7). An MRI scan will provide an estimate of soft tissue involvement. If there is concern about a vascular element to the lesion an angiogram is performed. This provides valuable information and helps in decision-making regarding presurgical selective embolization.
Chapter 42: Fibro-Osseous Lesions of the Skull Base
605
(A)
(B)
Figure 8 (A) Significant right cheek, eyelids, and temporal neurofibroma. (B) This was resected with preservation of a seeing eye. Further lid reconstruction will be performed.
When surgery is being considered, as mentioned above, it is of great value to separate the orbitotemporal lesions into three groups in terms of severity. As mentioned above, this is very helpful in determining which surgical procedure is indicated in terms of approach and magnitude (20).
Group 1 Minimal bony deformity—eye with useful vision with superior orbital fissure widened, displaced but in satisfactory position [Fig. 8(A)].
In addition to the bony deformity the upper eyelid is involved with resulting ptosis. Due to increased soft tissue bulk, the cheek soft tissue may also be affected. An MRI scan will provide the extent of soft tissue involvement and the CT scan illustrates the orbital and temporal deformities. In such a case using a coronal flap approach, the orbitofrontal area is exposed subperiosteally in addition to the temporal area. It is probably best to be conservative in such a case and simply remove as much neurofibroma as possible and contour the bone as required. It should be understood by both surgeon
606
Jackson
and patient that recurrence in the long term is likely after this form of treatment. This does not happen in every case and it is this fact that makes this somewhat conservative approach worthwhile initially. As much of the neurofibroma is removed as possible, but this is difficult.
Treatment There are various options available. In minor conditions with no functional problems observation is advised but this must be frequent with a treatment plan organized. This can then be instituted if, and when, necessary. Should the lesion be more severe then surgery should be considered but only by someone experienced in this area since both appearance and function have to be considered. The operative plan is to excise as much of the lesion as is safely possible, to reduce the orbital volume and to correct the eyelid position. A 3-D CT scan is mandatory since it will give an accurate appearance of the orbital deformity and will show if there is a posterior orbital wall defect. The presence of a significant defect may then necessitate neurosurgical help. The operation should be planned and executed in such a way as to establish symmetry. This will involve a coronal flap exposure, a vertical osteotomy of the lateral orbital wall continuing horizontally across the anterior aspect of the maxilla under the optic nerve and ending medially. A superior osteotomy on the lateral wall prepares the segment to be mobilized; if possible the floor of the orbit is included in this segment. The required amount of lateral wall is resected superiorly and the segment is moved upwards and stabilization is achieved superiorly with a plate. The resulting horizontal defect below the optic nerve is then grafted with cranial bone prior to its insertion. A small plate is placed on the graft, it is then bent over the infraorbital rim and stabilized with an anterior screw. It there is a defect in the orbital floor this is repaired with a split thickness cranial bone graft of the required dimensions. In these cases it is unusual to have to remove a significant amount of soft tissue, should this be necessary it is performed with careful preservation of nerves and muscles. The coronal incision is closed. The upper eyelid is now approached through a blepharoplasty incision; if there is ptosis the levator is exposed and shortened as in standard ptosis correction. If neurofibromatous tissue is present this should be resected as thoroughly as possible. Finally the excess eyelid skin is resected. In the lower eyelid it may be necessary to remove skin, muscle, and neurofibroma but often the most significant requirement is to transversely shorten the lid using a full-thickness wedge excision followed by a canthopexy. This will usually provide an acceptable end result (Fig. 8B).
Group 2 Orbit enlarged, temporal bone expansion occasionally with erosion, superior orbital fissure enlarged, eye displaced, no visual upset [Fig. 9(A)]. In this category the orbital soft tissue is more significantly involved—again there is subcutaneous neurofibroma, excess skin and ptosis of the upper lid. The lower lid is usually much less involved. The temporal fossa is enlarged and there may be characteristic thinning of the temporal bone due, as mentioned previously, to pressure of the temporal lobe gyri. The orbit is enlarged and is egg-shaped; the superior fissure is enlarged to a greater or lesser degree. The eye may be displaced but there are no visual problems. Diplopia is unusual due to the slow progression of the condition, which allows the eyes to accommodate. In some cases the maxilla may be somewhat underdeveloped and there can be intracranial involvement, especially in the frontotemporal area. A 3-D CT
and an MRI scan will give all the information necessary to plan the necessary surgical correction of both hard and soft tissues.
Treatment The soft tissue correction involves a direct transverse incision through the upper lid to remove the lesion [Fig. 9(B)]. Following this the lid ptosis is repaired together with adequate muscle and lid shortening. The orbital deformity is approached using a coronal flap and subperiosteal degloving of the frontal area, the orbit, and its contents. A frontosupraorbital osteotomy is performed, if indicated. The eye position is corrected by the planned osteotomies, resection of soft tissue, ostectomies, and bone grafts [Fig. 9(C)]. Basically, this is a superior and medial movement of the floor and lateral wall with plate stabilization. In order to correct the maxillary hypoplasia a full thickness cranial bone graft is harvested from the lateral aspect of the skull. This is then split, contoured, and used to augment the malar region using 3 to 4 screws for stabilization. The defects caused by the segmental orbital repositioning are filled with skull bone [Fig. 9(D)]. The other portions of the graft are placed back into the donor site and are stabilized with plates. Why full thickness? The harvest can be done under vision either with a contouring drill or by burr holes and a craniotomy drill. There is little or no chance of dural tear and consequently no problems with brain injury. Taking a partial skull graft with an osteotome always presents the risk of fullthickness penetration and dural damage. Should the eye need to be elevated because of orbital floor damage or displacement, a larger skull graft is taken, shaped as necessary, and placed into the orbital floor again with plate and screw fixation, if possible. The end result should be that of the eye being in its correct three-dimensional position [Fig. 9(E)].
Group 3 This is a much more severe and significant deformity, the correction of which needs experience and extremely careful planning. The eye is blind with a significant absence of the posterior wall of the orbit. As a result of this the eye is proptotic and pulsates. The proptosis is made more severe since these eyes are buthalmic and there is also an excess of intra-orbital soft tissue due to infiltration with neurofibroma [Fig. 10(A)]. There is severe upper lid ptosis, and the temporal area and the cheek soft tissue are increased in volume due to infiltration with the neurofibroma. The skin is discolored, somewhat scarred, and has irregularities resulting from the underlying lesion. The orbit is very enlarged and, again, it is egg-shaped [Fig. 10(B)]. There may be an arachnoid cyst present and the temporal lobe herniates into the orbit—this is the reason for the extreme proptosis and pulsations. There is temporal bone expansion, absence of the posterior wall of the orbit, the eye is blind, and there is upper lid ptosis. The eyelids are considerably expanded in all dimensions. The bone in the temporal region is expanded laterally and is frequently thinned and eroded by the cerebral gyri in many areas.
Treatment Once again, the approach is by a coronal flap and by orbital exenteration. Since the eye is blind and greatly enlarged with frequent distortion of its internal anatomy, it should be removed together with all of the orbital contents. The internal lamella, the lid margins, and the eyelashes are removed leaving only the external eyelid skin. The fronto-orbital region is approached through a bicoronal incision. In some cases the exenteration will be completed from above. In order to
Chapter 42: Fibro-Osseous Lesions of the Skull Base
607
(A)
(B)
Figure 9 (A) Neurofibromatosis involving the right upper eyelid. (B) Resection of upper eyelid skin excess and removal of neurofibroma. (C) Planned resection of lateral wall of egg-shaped orbit. (D) Osteotomies and bone grafting to correct the shape of the orbit. (E) Postoperative appearance. (C–E shown on page 602).
reduce the enlarged orbit osteotomies as performed on the lateral orbital wall, the medial orbital wall, the zygomatic arch, and transversely below the infraorbital rim [Fig. 10(C)]. This is completed taking care to preserve the infra-orbital nerve. This allows the orbital dimensions to be reduced vertically and transversely with a corresponding decrease in orbital volume. This can be performed accurately with good
preosteotomy planning. Microplates and screws provide a stable reconstruction but wires can be used if preferred. The large defect, which is always present in the posterior orbital wall is bone grafted using ribs or cranium; once again rigid fixation should be achieved. The superior orbital rim is usually enlarged cranially and this should also be bone grafted on its orbital aspect in order to achieve orbital dimensions to
608
Jackson
(C)
(E)
(D)
Figure 9 (Continued)
equal those of the normal orbit. It is also necessary to bone graft the horizontal maxillary defect. Osseo-integrated implants are placed for future reconstruction with an external framework. The well-vascularized thin upper and lower eyelid skin is now used to re-surface the orbit, this is the key to success and is made possible by the slowly growing neurofibroma having acted as a tissue expander of the lids. After elevation of the lid skin, all tissues deep to this are sacrificed, i.e. the conjunctiva and tarsal plates. Laterally the excess expanded temporal skin is excised and closed. Once this has
been done the orbit is packed in order to ensure that the skin adheres to the underlying bony orbital reconstruction. On some occasions, e.g. when bone is exposed and there is a soft tissue defect, free tissue transfer may be used [Fig. 10(D)–10(F)]. With later contouring a very acceptable result can be achieved [Fig. 10(G) and 10(H)]. On one occasion there was so much erosion of the lateral temporal bone by pressure of the expanded temporal lobe that a large titanium plate was placed as an onlay chosen to reconstruct and reinforce that total area. It was felt that this
Chapter 42: Fibro-Osseous Lesions of the Skull Base
(A)
609
(B)
(C)
Figure 10 (A) Neurofibroma of left cheek and orbit. There is excessive skin and the eye has been removed elsewhere. (B) On exploration the orbit is typically egg-shaped and the posterior wall is partially absent. (C) Osteotomies and bone grafting (skull grafts) to correct orbital shape and rib grafts to the posterior orbital defect. (D–F) Rectus abdominis free flap to reconstruct skin and subcutaneous tissue. (G, H) Pre- and postoperative results with an osseointegrated prosthesis. (D–H shown on page 604.)
would give a more stable situation over the long term. This has been the case on a 6-year follow-up [Fig. 11(A)–11(C)]. In the long term, when the area has been stabilized totally, and this is a clinical decision, the orbital osseointegrated implants are exposed, an external framework is placed, and this is used to hold a composite eye and eyelid prosthesis in position. Without a stable deep orbit the volume of the prosthesis cannot be accommodated. The edge of the prosthesis is arranged in such a way that the rim of the spectacles that we advise the patient to wear lies directly over it. With a good prosthesis the reconstruction can be excellent [Fig. 12(A)–12(C)]. It also allows a good field for accurate follow-up. On some occasions the dimensions of the primary
soft tissue defect will be such that a free tissue transfer with later modification will be necessary. This situation provides a much greater challenge in reconstruction. It also takes longer because it must be staged. In some cases there is a widespread involvement of the temporo-orbital cheek area but once again the tumor has acted as a tissue expander. The resection as described above is carried out, it is considerably more extensive but, in our experience, due to the skin expansion mentioned above, skin cover can usually be achieved. The orbit and its contents are managed as described earlier as is the reconstruction. If there is extensive skin involvement, and due to the constraints insurance companies put on the patients as far as how long they are allowed to stay in hospital, the
610
Jackson
(E)
(F) (D)
(G)
(H)
Figure 10 (Continued)
resurfacing is not always as good as it could be. If it were possible to carry out more surgery, skin expansion would be used to gain a superior result since all of the involved skin and soft tissue could then be resected, the area would then be covered with the expanded skin. In spite of this a great majority of our patients have been satisfied. Patients coming from abroad and seeking treatment require special mention. They form about 20% of our cases, and require as rapid a reconstruction as possible (Fig. 13).
BILATERAL FACIAL NEUROFIBROMATOSIS This is a rare condition and, once again, multiple stages are required. The approach is that described for the unilateral cases but the length of treatment is much more prolonged (Fig. 14).
Complications The complications have been mainly related to bleeding and this can be massive (Table 1). Thus, in large lesions pre-
operative angiography is advised and, if possible and indicated, embolization should be performed. In 118 surgical interventions, there were 9.6% minor complications and 6% major complications—all hematomas. In severe intraoperative bleeds argon laser was used on one occasion. In another, muscle was harvested and beaten in order to release clotting agents. It was packed into the bleeding area, and this provided successful hemostasis. In three other cases ribbon gauge packing was employed with delayed closure after cautious, staged removal of the gauze. It should be noted that we prefer not to use free tissue transfer, if possible. It is not usually considered because the neurofibromatosis acts as a natural tissue expander, the skin is well vascularized and flaps can be constructed and thinned as required. These are challenging problems but with advances in radiology it is now possible to plan the required skeletal adjustments. The experience gained in congenital skull base surgery makes for ease in exposure, osteotomies for exposure, resection, and reconstruction. Skull bone grafts, free tissue transfer, and plate and screw fixation have reduced
Chapter 42: Fibro-Osseous Lesions of the Skull Base
611
(A)
(B)
(C)
Figure 11 Right temporal expansion. (A) The orbit is greatly enlarged, mainly downwards. There is significant expansion of the lateral temporal fossa with many bony defects. (B) After correction—reinforced with heavy metal mesh. (C) Soft tissue reconstruction using a rectus abdominis free tissue transfer.
612
Jackson
(B)
(A)
(C)
Figure 12 (A) Significant involvement and enlargement of the right orbit. (B) Resection of neurofibroma, reduction of right orbit with osteotomies and bone grafting of posterior wall defect. Reconstruction with eyelid skin. (C) Reconstruction with an osseointegrated implant. This is much improved when the patient wears spectacles.
Chapter 42: Fibro-Osseous Lesions of the Skull Base
613
(A)
(B)
Figure 13 (A, B) A young man from Africa with significant soft tissue and skeletal involvement. Pre- and postoperative result following soft tissue resection, removal of orbital contents, orbital reduction by osteotomies, and osseointegrated implants.
operating time and have increased accuracy of reconstruction and prevention of complications. Cooperation between plastic surgeon, neurosurgeon, and sometimes oral surgeon has increased efficiency and safety. This is no longer an area where surgeons fear to tread (21,22). Neurofibromatosis has one redeeming feature— patients with this disease cannot produce hypertrophic scars no matter the size of the defect, the condition of the wound edges, or infection. Unfortunately, coupled with this is the abnormal stretchability of the skin.
WEGENER GRANULOMATOSIS This unusual condition usually presents after many resections and recurrences. The usual problem is that the treatment has not been sufficiently radical. This condition should be approached almost as though it is a malignancy. A block resection should be performed no matter what tissue is involved. It may require a skull base, an orbital, or a nasal resection. In the fronto-orbital area a combined facial and coronal approach may be necessary to resect any nasal, frontal,
614
Jackson
Figure 14 Bilateral neurofibromatosis patient is blind in both eyes.
or orbital bone and again the orbital contents may have to be sacrificed. The best cover is by a galeofrontalis flap. This is well vascularized and can be packed into defects; it can also provide orbital resurfacing if necessary. This results in an excellent vascularized surface that will accept a split thickness skin graft if necessary. In larger defects, it may be necessary to use a free muscle flap repair, again chosen and tailored to the required volume. The esthetic results often leave a lot to be desired and, once again, a prosthesis may be required for rehabilitation. REFERENCES 1. Gross CW, Montgomery WW. Fibrous dysplasia and malignant degeneration. Arch Otolaryngol. 1967;85:653–657. 2. Huvos AG, Higinbotham NL, Miller TR. Bone sarcoma arising in fibrous dysplasia. J Bone Joint Surg Am. 1972;54:1047. 3. Posnick JC, Hughes CA, Milmoe G, et al. Polyostotic fibrous dysplasia: An unusual presentation in childhood. J Oral Maxillofac Surg. 1996;54(12):1458–1464. 4. Jackson IT, Hide TAH, Gomuwka PK, et al. Treatment of cranioorbital fibrous dysplasia. J Maxillofac Surg. 1982;10:138–141. 5. Moore AT, Buncic JR, Munro IR. Fibrous dysplasia of the orbit in childhood. Clinical features and management. Ophthalmology. 1985;92:12–20. 6. Munro IR, Chen YR. Radical treatment for fronto-orbital fibrous dysplasia: The chain-link principle. Plast Reconstr Surg. 1981;67:719–730. 7. Munro IR. Treatment of craniomaxillofacial fibrous dysplasia: How early and how extensive? (Discussion). Plast Reconstr Surg. 1990;86:843–844. 8. Chen Y-R, Noordhoff MS. Treatment of craniomaxillofacial fibrous dysplasia: How early and how extensive? Plast Reconstr Surg. 1991;87(4):799–800. 9. Ragab MA, Mathog RH. Surgery of massive fibrous dysplasia and osteoma of the midface. Head Neck Surg. 1987;9:202–210.
10. Posnick JC. Fibrous dysplasia of the craniomaxillofacial region: Current clinical perspectives. Br J Oral Maxillofac Surg. 1998;36:264–273. 11. Middleton DS. The place of surgery in fibrous dysplasia of the jaws. J R Coll Surg (Edin.). 1963;8:310–318. 12. Posnick JC, Wells MD, Buncic JR, et al. Childhood fibrous dysplasia presented as blindness: A skull base approach for resection and immediate reconstruction. Pediatr Neurosurg. 1993;19(5):260–266. 13. Schwartz DT, Alpert M. The malignant transformation of fibrous dysplasia. Am J Med Sci. 1964;246:1–20. 14. Slow IN, Friedman EW. Osteogenic sarcoma arising from a pre-existing fibrous dysplasia. Report on a case. J Oral Surg. 1971;29(2):126–129. 15. Tanner HC, Dahlin DC, Childs DS. Sarcoma complicating fibrous dysplasia: Probable role of radiation therapy. Oral Surg Oral Med Oral Pathol. 1961;14:837–846. 16. Batsakis JG. Tumors of the peripheral nervous system. In: Batsakis JG, ed. Tumors of the Head and Neck: Clinical and Pathological Considerations. 2nd edn. Baltimore: William & Wilkins, 1979:313–333. 17. Poyhonen M, Niemela S, Herva R. Risk of malignancy and death in neurofibromatosis. Arch Pathol Lab Med. 1997;121:139– 143. 18. Krastinova-Lolov D, Hamza F. The surgical management of cranio-orbital neurofibromatosis. Ann Plast Surg. 1996;36:263– 269. 19. Jackson IT, Carbonell A, Potparic Z, et al. Orbitotemporal neurofibromatosis: Classification and treatment. Plast Reconstr Surg. 1993;92:1–11. 20. Jackson IT, Laws ER, Jr., Martin RD. The surgical management of orbital neurofibromatosis. Plast Reconstr Surg. 1983;71:751– 758. 21. Marchac D. Intracranial enlargement of the orbital cavity and palpebral remodeling for orbit-palpebral neurofibromatosis. Plast Reconstr Surg. 1984;73:534–543. 22. Van der Meulen JC, Moscona AR, Vandrachen M, Hirshowitz B. Orbital neurofibromatosis. Ann Plast Surg. 1982;8(3):213–220.
43 Metastatic Skull Base Tumors Krishna Satyan and Sujit S. Prabhu
ation of a proper work-up. These pain syndromes are usually caused by invasion of the bone, and can often precede neurologic findings by weeks. The clinical manifestation of these lesions depends on the site of involvement and the rate of growth. Additionally, even when the diagnosis is suspected, the anatomic location of the lesion is still often uncertain, as different sized lesions may cause similar clinical presentations, depending on what anatomic structures are involved. Five clinical syndromes have been associated with skull base metastasis—orbital, parasellar, middle-fossa, jugular foramen, and occipital condyle syndromes. In the recent review by Laigle-Donadey et al., the most common syndrome was parasellar/sellar, accounting for 29% of patients. Middle fossa and jugular foramen syndromes were the least common, accounting for 6% and 3.5% respectively. Of note, in 33% of cases a specific syndrome could not be identified. An uncommon but highly interesting presentation is the hemibasis syndrome, where the entire base of the skull may be involved. This syndrome is characterized by progressive ipsilateral paralysis of at least seven cranial nerves, without any long-tract signs or raised intracranial pressure.
Although the skull is a common site of metastases, involvement by metastatic disease of the ethmoid, sphenoid, basal aspects of the frontal, temporal, and occipital bones is rare. Partly, this is because skull base lesions are inconsistent in their clinical behavior and diagnosis is difficult (1). Metastases to the skull base from distant tumors occur in approximately 4% of cancer patients (1). However, autopsy reports have shown the incidence to be higher, especially in specific anatomic locations. Belal discovered a 3% incidence of temporal bone metastases in the general population itself (2), while Jung et al. reported an approximately 24% incidence of temporal bone metastases in patients with a history of malignancy (3). Almost all cancers may spread to the skull base. The most common malignancies to metastasize to the skull base include breast, lung, and prostate cancer. Breast cancer is the most common etiology in women, and prostate cancer in men. In a recent review of 279 cases of skull base metastases, prostate cancer accounted for 38.5% of patients, while breast cancer accounted for 20.5% (Table 1). Dissemination to the skull base generally occurs late in the disease process and hence the incidence of clinically significant lesions is low. At the time of diagnosis anywhere from 20% to 100% of patients already have widespread metastases, including other bony sites. However, skull base lesions can be the first presenting sign of cancer, as was the case for 28% of patients in a recent review of the literature (1). In approximately 50% to 65% of cases, paranasal sinus metastasis may be symptomatic before the primary tumor is identified (4). Renal cell cancer is the most common primary tumor to spread to the paranasal sinuses, accounting for 40 of 69 patients in one study (5). The maxillary sinus is the most common sinus affected (50%), followed by the ethmoid sinus, frontal sinus, nasal cavity, and sphenoid sinus (4). The majority of skull base metastases probably occur via direct hematogenous spread. This may occur from the initial primary tumor, or via a secondary site such as lung metastasis. A second proposed mechanism of spread is through Batson’s valveless venous plexus, which connects pelvic structures to the skull. This latter mechanism may represent the pathophysiology behind the seeding of primary tumors such as prostate cancer and other retroperitoneal tumors (1). Increased intra-abdominal and intra-thoracic pressure can shunt blood through Batson plexus to the basilar plexus of veins (4). The vast majority of skull base metastases are clinically silent, secondary to their occurrence late in the disease process. They may become symptomatic when their growth produces pain or cranial nerve palsies, caused by osseous invasion and invasion or compression of the cranial nerves as they exit the basal foramina. The occurrence of craniofacial pain in patients with a history of cancer should alert clinicians to the possibility of skull base metastasis and prompt the initi-
ORBITAL SYNDROME The orbital syndrome occurs in 7% to 12.5% of skull base metastases, and is a rare presentation of orbital tumor. It is an unusual complication of systemic cancer, and rarely isolated. The orbital syndrome is characterized by frontal headaches, consisting of dull, persistent pain over the affected eye. It is associated with diplopia, and often preceded by blurred binocular vision. It can also be associated with numbness in the frontal region. Patients usually have proptosis and external ophthalmoplegia of the involved eye. The syndrome may be associated with sensory loss in the distribution of the ophthalmic segment of the trigeminal nerve. Occasionally, the tumor is felt when the orbit is palpated, and periorbital swelling may be seen. Decreased vision due to optic nerve compression has been reported. The most common metastases associated with this syndrome are prostate cancer (56%), lymphoma (23%), and breast cancer (15%) (1). In Fig. 1(A) and 1(B) the patient presented with new onset right orbital pain and proptosis. The tumor involving primarily the greater wing of the sphenoid and protruding into the orbit was the initial presentation of a case of follicular carcinoma of the thyroid. The patient had a gross total resection of the tumor.
SELLAR/PARASELLAR SYNDROME The sellar/parasellar syndrome is the most common presentation of skull base metastases, representing 29% of cases in the review by Laigle-Donadey et al. Because of the close 615
616
Satyan and Prabhu
Table 1 Characteristics of Skull Base Metastases in a Literature Review of 279 Patientsa Main types of cancer (Number of cases = 279)
Prostate (38.5%) Breast (20.5%) Lymphoma (8%) Lung (6%) Others (27%) Orbit (12.5%) Parasellar and sellar (29%) Gasserian ganglion (6%) Jugular foramen (3.5%) Occipital condyle (16%) Others (33%)
Main clinical syndromes (Number of cases = 275)
a
Full data were not available for all patients.
proximity and overlying symptomatology, lesions in the parasellar (cavernous sinus) and sellar regions are grouped together. Metastases to the parasellar regions actually account for 7% to 16% of skull base metastases. Tumors of the parasellar region tend to affect the ocular motility nerves and the divisions of the trigeminal nerve as they course through or near the cavernous sinus. Patients will have unilateral supraorbital frontal headache and palsies of the oculomotor, abducens, and trochlear nerves. Unlike the orbital syndrome, there is no sign of proptosis. Vision is usually not impaired, unless the disease is at a very late stage. Given the involvement of the trigeminal nerve (usually ophthalmic division and occasionally maxillary division), patients can also present with facial paresthesias and pain. Lymphoma has a particular affinity for the cavernous sinus, and metastases may present with visual loss as the initial manifestation of the systemic disease (1). Metastatic tumors to the pituitary gland are found at only 1% of all pituitary tumor resections. Most pituitary metastases are asymptomatic, with only 7% reported to be symptomatic. Diabetes insipidus, anterior pituitary dysfunction, visual field defects, headache/pain and opthalmoplegia are the most commonly reported symptoms. Of these symptoms, diabetes insipidus is especially common in this population, occurring between 29% and 71% of patients who experience symptoms (6). Of the metastases, breast and lung cancers are the most common diseases that metastasize to the pituitary gland. Certain neuroimaging features such as thick-
(A)
(B)
Figure 2 (A) A coronal T1-weighted MRI scan following gadolinium administration showing metastasis in the right Meckel cave. (B) Coronal MRI scan after administration of gadolinium 6 months later on the same patient as Figure 2(A) showing stable changes in Meckel cave following treatment with stereotactic radiosurgery. There is a new metastasis in the pituitary fossa with right optic chiasmal involvement. Patient underwent trans-sphenoidal resection of this tumor.
ening of the pituitary stalk, invasion of the cavernous sinus, and sclerosis of the surrounding sella turcica may indicate metastasis to the pituitary gland. Also, differentiating metastasis to the pituitary gland from bone metastasis to the skull base in which the sella turcica is involved can also be difficult. Since a vast majority of cases of pituitary metastases occur in association with multiple systemic metastases, they are usually associated with end-stage disease. In an autopsy series, Kovacs found metastatic lesions in other organs in all 18 patients with pituitary metastasis (7). Mean survival rates have been reported to be between 6 and 22 months, independent of the treatment strategy. Surgery, radiation, and chemotherapy form the mainstay of a multimodality treatment for these tumors. The patient in Figure 2(B) presented with decreased visual acuity of the right eye. The magnetic resonance image (MRI) confirmed a metastasis to the pituitary gland. At the time of diagnosis this represented the only site of active disease of the patient’s lung primary and, because of the accompanying symptoms, the patient had a trans-sphenoidal resection of the mass with improvement in symptoms.
MIDDLE FOSSA SYNDROME
(A)
(B)
Figure 1 (A) Axial MRI scan following gadolinium administration showing a metastasis from follicular carcinoma of the thyroid to the right sphenoid wing and the lateral wall of the orbit causing proptosis and right orbital pain. (B) Coronal MRI scan after gadolinium administration showing the same patient scans as in Figure 1(A) with encroachment of the tumor into the orbit. This tumor was radically resected.
This syndrome is also called the Gasserian ganglion syndrome, and is characterized by facial parasthesias, numbness, and pain. Symptoms are usually in the trigeminal distribution, but tend to spare the frontal region. Headache is unusual, and patients often complain of a dull, achy pain in the cheek, jaw, or forehead. Symptoms occasionally may mimic trigeminal neuralgia, with lightning-like spasmodic pain (1,4). Numbness usually starts close to the midline, progressing laterally to the anterior part of the ear (4). Patients usually describe sensory loss in the maxillary and mandibular distributions, and occasionally patients will complain of numbness in the ophthalmic distribution (1). Motor root dysfunction is also occasionally seen. As the tumor spreads beyond the Gasserian ganglion, the nerves of extraocular motility and facial nerve may become involved. In the series by Greenberg et al., 4 of 15 patients with middle fossa syndrome had sixth nerve palsy, 4 had combined extraocular palsies, and 3 had seventh nerve palsy (8). The patient in Figure 2(A)
Chapter 43: Metastatic Skull Base Tumors
presented with symptoms of right facial numbness and a metastatic tumor from a primary lung carcinoma was diagnosed. The patient subsequently underwent stereotactic radiosurgery with improvement in symptoms and a significant reduction in the size of the mass at 6 months.
JUGULAR FORAMEN SYNDROME In the review of 43 patients with metastases to the skull, Greenberg et al. described a syndrome primarily related to paralysis involving the lower cranial nerves with evidence of glossopharyngeal, vagus, and accessory nerve dysfunction. Hoarseness and dysphagia were the common presenting symptoms (8). Clinical examination reveals paralysis with amyotrophy of the ninth through eleventh cranial nerves, affecting the palate, vocal chord, ipsilateral sternocleidomastoid, and upper part of the trapezius muscles sometimes associated with a Horner syndrome. Pain can also be an associated symptom with glossopharyngeal neuralgia, and unilateral pain usually behind the ear on the involved side (9).
OCCIPITAL CONDYLE SYNDROME This syndrome usually consists of unilateral pain in the occipital region followed a few weeks later by ipsilateral twelfthcranial nerve palsy. This clinical presentation may be the first sign of metastatic disease (10,11). On examination, the patients typically hold their neck rigidly and there is tenderness of the scalp in the affected occipital region, with the ipsilateral tongue being weak, atrophic, and sometimes exhibits fasciculations.
TEMPORAL BONE INVOLVEMENT Approximately 20% of patients may have metastasis to their temporal bones. The most common tumor to spread to the temporal bone is breast cancer, but a recent study showed that the ratio of males to females was close to unity. Conductive hearing loss is the most common symptom in these patients (40%), but patients may present with vertigo, facial palsy, or tinnitus, amongst other symptoms. A relatively high percentage of patients (36%) with temporal bone metastasis may be asymptomatic. The petrous apex is the most commonly involved site, while the osseous and membranous labyrinths are rarely affected (12).
DIAGNOSIS OF SKULL BASE METASTASES Cranial neuropathy in a patient with known primary disease should raise a strong suspicion for skull base metastasis. However, other malignant disorders can cause similar clinical findings, and must be distinguished. Leptomeningeal carcinomatosis and spread to the brainstem or cerebellum can mimic metastatic disease involving the skull base (8). Magnetic resonance imaging (MRI) sequences before and after intravenous gadolinium administration is the best method to detect skull base metastases and to identify the soft tissue involvement. Fat suppression techniques with gadolinium infusion are particularly useful. Osseous metastasis usually is seen as a hypointense lesion on a non-enhanced T1weighted image with substitution of the usual hyperintense fat signal. The T1-weighted sequences with fat suppression
617
generally show variable enhancement after gadolinium administration. MRI imaging is also useful to show metastasis in the dura of the skull base particularly in relation to the cavernous sinus and Meckel cave regions [Fig. 2(A)], as well as the cranial nerves and leptomeninges (4). High definition skull base CT scans with reconstructions are also very useful in showing the bony involvement of the skull base by the metastasis and the various foramina involved. These sequences can be particularly useful when planning surgical approaches. The lesions in skull base can be either lytic or sclerotic lesions, with most disease processes producing a lytic picture. In prostate and lung cancers, a blastic or mixed pattern is more commonly seen. Radionuclide bone scans are extremely sensitive in detecting bone metastases. PET (positron emission tomography) scans are also useful in demonstrating skull base lesions, especially when CT and MRI are inconclusive. About 30% to 50% of lesions seen on bone scans are not detected by CT or MRI. Bone scans may also show increased activity with inflammatory conditions such as sinusitis, mastoiditis, or temporomandibular joint arthrosis, potentially leading to misdiagnosis. However, bone single photon emission computed tomography (SPECT) now allows for better discrimination between inflammatory conditions and cancer when compared to planar bone scans (4). Most skull base metastases are not readily accessible for purposes of a biopsy because of their relationship to critical structures in the skull base. However, biopsies may be performed in accessible lesions of the anterior cavernous sinus either by orbital fine-needle aspiration biopsy or by a transsphenoidal approach especially to rule out nonmetastatic processes like infections. Cerebrospinal fluid examination is of great value to rule out infection or leptomeningeal carcinomatosis, which may accompany skull base metastasis. The index of suspicion should be high especially in metastases involving the basal dura.
PROGNOSIS Skull base metastasis is often a late event in the course of the primary cancer and death generally is due to the systemic progression of the disease. In a review of skull base metastases by Leigle-Donadey et al., the overall median survival was 31 months (1). In this review, breast carcinoma was associated with the best survival (60 months) whereas lung and colon carcinomas were the poorest (2.5 and 1.5 months respectively). Prostate carcinoma and lymphoma showed intermediate survival of 21 and 24 months respectively. Cranial palsies indicate a poor prognosis with an average survival of 25 months (from a few days to 20 months) after the onset of cranial involvement. In a review of cranial deficits in patients with metastatic prostate cancer, McDermott et al. showed that 10 of the 15 patients (67%) died within 3 months of developing symptoms, and the remaining 5 patients lived 9 months and 31 months from the onset of symptoms (13).
TREATMENT Radiotherapy as a stand-alone treatment or in combination with chemotherapy or with surgery is the standard treatment of skull base metastases. Stereotactic radiosurgery can be used in selected cases of metastases involving the sella and the parasellar regions. In a recent review, Iwai et al. treated and followed up 21 patients with cavernous sinus metastases
618
Satyan and Prabhu
and local invasion of nasopharyngeal carcinomas to the cavernous sinus using gamma knife radiosurgery (14). Nine of these patients had nasopharyngeal cancer, and 12 had distant metastases from other cancers. The radiation dose to the tumor margin was 10 to 21 (median 14) Gy. The actual 1and 2-year tumor growth control rates were 68% and 43% respectively. The mean survival time was 13 months with no patients having any radiation injury. They concluded that gamma knife radiosurgery was a useful therapeutic option for the treatment of cavernous sinus metastases and invasion. This modality of treatment is well tolerated by patients even in those with widespread metastatic disease. Chemotherapy and hormonal therapy are most often used in combination with radiation therapy (RT) and may produce clinical improvement with survival longer than 2 years in some cases of prostate and breast carcinoma. Surgical resection of skull base metastases is usually reserved for patients with single metastasis. In the region of the sella and parasellar regions the indications for surgery would be to obtain diagnostic tissue and in cases of worsening neurological function [optic apparatus compression in sellar lesions, see Fig. 2(B)] to decompress neural tissue. In a review of metastases to the pituitary gland, Fasset and Couldwell noted that gross total resection was difficult for a number of reasons including vascularity of the tumor, resulting in heavy bleeding, local invasion, or invasiveness into the surrounding cavernous sinus and infiltration of the hypothalamus and optic nerves (6). Reports on two surgical series have indicated no difference in survival attributed to resection. A review of 36 patients by Morita et al. found no statistically significant difference in survival in the 21 patients who underwent surgery. However, they found an improvement in symptoms (visual acuity, pain, and ophthalmoplegia) and quality of life with an aggressive tumor resection and radiation therapy. In another review, Branch and Laws showed no survival benefits associated with surgery but confirmed improved quality of life in this patient population (15).
CONCLUSIONS Skull base metastases are an uncommon presentation of metastatic disease. Pain and cranial nerve involvement are usually the presenting symptoms and one should have a high index of suspicion of skull base metastasis in this situation.
The overall survival of patients with multimodality treatment is still modest, and this may be a reflection of the widespread systemic disease of the patient at the time of involvement of the skull base. Early surgical intervention may potentially improve quality of life in a select group of patients.
REFERENCES 1. Laigle-Donadey F, Taillibert S, Martin-Duverneuil N, et al. Skullbase metastases. J Neuro-Oncol. 2005;75(1):63–69. 2. Belal A, Jr. Metastatic tumors of the temporal bone. A histopathological report. J Laryngol Otol. 1985;99:839–846. 3. Jung TT, Jun BH, Shea D, et al. Primary and secondary tumors of the facial nerve. A temporal bone study. Arch Otolaryngol Head Neck Surg. 1986;112:1269–1273. 4. DeMonte F, Hanabali F, Ballo MT. Skull base metastasis. Textbook of Neuro-Oncology Philadelphia, PA: Elsevier Saunders, 2005:466–475. 5. Bernstein JM, Montgomery WW, Balogh K, Jr. Metastatic tumors to the maxilla, nose, and paranasal sinuses. Laryngoscope. 1966;76:621–650. 6. Fassett DR, Couldwell WT. Metastases to the pituitary gland. Neurosurg Focus. 2004;16(4):E8. 7. Kovacs K. Metastatic cancer of the pituitary gland. Oncology. 1973;27(6):533–542. 8. Greenberg HS, Deck MD, Vikram B, et al. Metastasis to the base of the skull: Clinical findings in 43 patients. Neurology. 1981;31(5):530–537. 9. Svien HJ, Baker HL, Rivers MH. Jugular foramen syndrome and allied syndromes. Neurology. 1963;13:797–809. 10. Capobianco DJ, Brazis PW, Rubino FA, et al. Occipital condyle syndrome. Headache. 2002;42(2):142–146. 11. Moeller JJ, Shivakumar S, Davis M, et al. Occipital condyle syndrome as the first sign of metastatic cancer. Can J Neurol Sci. 2007;34(4):456–459. 12. Gloria-Cruz TI, Schachern PA, Paparella MM, et al. Metastases to temporal bones from primary nonsystemic malignant neoplasms. Arch Otolaryngol Head Neck Surg. 2000;126:209– 214. 13. McDermott RS, Anderson PR, Greenberg RE, et al. Cranial nerve deficits in patients with metastatic prostate carcinoma: Clinical features and treatment outcomes. Cancer. 2004;101(7):1639–1643. 14. Iwai Y, Yamanaka K, Yoshimura M. Gamma knife radiosurgery for cavernous sinus metastases and invasion. Surg Neurol. 2005;64(5):406–410; discussion 410. 15. Branch CL, Jr., Laws ER, Jr. Metastatic tumors of the sella turcica masquerading as primary pituitary tumors. J Clin Endocrinol Metabol. 1987;65(3):469–474.
Index
Page number with f indicates figure and t indicates table. ACC. See Adenoid cystic carcinoma Accessory nerve, 403 Acoustic neuroma, 107–109, 110, 110t Acoustic reflex testing, 107 Acoustic schwannoma, 163–164, 194 Acromegaly, 561f ACTH. See Adrenocorticotropic hormone Adamantinomatous craniopharyngioma, 574, 579 Adenocarcinoma clinical features, 449 outcome and prognosis, 450–451 pathology, incidence, and epidemiology, 448–449 treatment, 449–450 Adenoid cystic carcinoma (ACC), 70, 88f , 90f , 177, 241f , 262f , 268, 295, 317 clinical features, 447 incidence and epidemiology, 446–447 outcome and prognosis, 447–448 pathology, 447 treatment, 447 Adenomatous polyposis coli (APC), 481 Adjuvant therapy, 294, 508–509 Adjuvant therapy for craniopharyngioma chemotherapy, 579 external beam radiation, 579 gamma knife, 579 radiation therapy, 579 Adrenalectomy, 567 Adrenocorticotropic hormone (ACTH), 161, 357, 557, 564, 568 Allergic fungal sinusitis, 45–46 Alveolar osteotomies, 150 Alveolar rhabdomyosarcoma (ARMS), 75 Amelanotic tumors, 461 Ameloblastoma, 58, 59f Anaplastic meningioma, 506f Anaplastic pituitary tumors, 560f Anatomic imaging techniques, 296 Anesthesia and intraoperative monitoring anesthesia and neuroprotection, 125–126 complications (skull base procedure) arrhythmias, 121–122 blood loss, 122 cerebrovascular complications, 121 macroglossia/facial swelling, 121 peripheral nerve/cranial nerve injuries, 122 pneumocephalus, 121 venous air embolism, 120–121 for endoscopic surgery, 126–127 monitoring and anesthesia, 122–125 positioning lateral position, 120 park-bench position, 119 prone position, 119 sitting position, 120 supine position, 119 postoperative management, 127–128
with vascular lesions, 125–126 Anesthetics on somatosensory evoked potentials, 124t Angiofibroma, 295 Angiofibromas and vascular tumors. See Vascular tumors of skull base Angiogram postoperative, 215f preoperative, 214f Angiography, 239, 485–486 Angiosarcoma, 53 Anosmia, 100 Anterior base of skull tumor (ABST) patients, 201 Anterior communicating artery (ACOM), 207 Anterior cranial fossa, 231 Anterior endocranial surface. See under Cranial base anatomy Anterior maxillary antrotomy, 280 Anterior nasal cavity, 269 Anterior petrosectomy, 287, 370, 372 Anterior rhinoscopy, 99 Anterior skull base, 182–183 Anterior skull base tumors clinical assessment, 236, 295 complications/avoidance, 300 cranial base reconstruction, 260–261 craniofacial resection approach, 250–256, 257f bony dissection and osteotomy, 297 incision, 296 preoperative preparation, 296 soft tissue dissection, 296–297 diagnostic imaging, 295–296 biopsy, 296 CT, 295 MRI, 295 PET Scan, 295–296 vascular encasement, 296 endoscopic surgical approaches, 246f , 298–300 midfacial degloving (MFD) approach, 297 orbitozygomatic approach, 297 pathology and diagnosis adenoid cystic carcinoma, 295 angiofibroma, 295 chondrosarcoma, 293–294 chordomas, 294 esthesioneuroblastoma, 293 lymphoma, 294–295 malignant melanoma, 294 meningioma, 293 nasopharyngeal carcinoma, 294 postoperative care, 300 preoperative preparation, 240–242 quality of life, 192–197, 295, 300 subcranial approach, 297 surgery, 242–244 surgical anatomy, 231, 293 surgical technique, 296 Anterior Inferior Cerebellar Artery (AICA), 40, 389, 410 Anterolateral approach, 288 Anthracyclines, 474–475 Antibody microarrays, 66
619
620
Index
Antineutrophil cytoplasmic antibodies (c–ANCA), 47 Antonio De La Cruz classification, 543t Antrum of Highmore, 228 Arachnoid cysts. See under Cysts of skull base Arrhythmias, 121–122 Arterial encasement. See under Cerebrovascular management Asymptomatic meningioma, 508 Attention-deficit hyperactivity disorder (ADHD), 514 Atypical meningioma, 506f Audiometry, 106–107 Auditory brainstem responses (ABR), 109, 110, 111t, 112 Axial bone window CT scan, 585f Axial MRI, 168f Axial diffusion-weighted MRI, 585f , 590f Bacillus Calmette–Guerin (BCG) vaccine, 468 Basiocciput, 309 Basisphenoid bone, 309 Benign tumors, 49, 159, 267 Bicoronal approaches, 251, 258 Bilateral carotid body tumors, 548–549 Bilateral facial neurofibromatosis, 610–613 Biopsy, 239–240, 296 Bitemporal hemianopsia, 557 Blood loss, 122 Blowout injury, 121 Bony dissection and osteotomies, 349–352 Brain stem auditory evoked potentials (BAEPs), 123 Bromocriptine, 566 Bruns nystagmus, 97 Bypass techniques. See under Cerebrovascular management Cabergoline, 566 Caloric testing, 112, 114 Carbon ion therapy, 168–169 Carcinoid tumors, 56 Carotid body paragangliomas, 542f , 548 Carotid body tumors (CBTs), 539, 540, 543, 543f Cavernous sinus, 308–309, 327 Cavernous sinus meningioma, 218–220 Cavitron ultrasonic aspirator, 299 Cell survival curve, 159, 160f Central nervous system (CNS), 503 Central skull base (CSB), 82, 84 Cerebellar mutism syndrome (CMS), 184 Cerebellopontine angle (CPA), 389, 583 Cerebellopontine angle tumors clinical assessment, 390–391 complications and avoidance CSF leak, 399 facial nerve dysfunction, 400 hydrocephalus, 399 meningitis, 399–400 postoperative hematoma, 399 vertigo, 400 diagnostic imaging computed tomography, 391 magnetic resonance imaging, 391–393 pathology and diagnosis, 389–390 postoperative care, 399 preoperative preparation, 393–394 surgical anatomy, 389 surgical technique retrosigmoid approach, 394–396 translabyrinthine approach, 396–399 Cerebral angiography, 207, 277 Cerebral vasospasm, 217 Cerebral venous sinuses, 212t, 215 Cerebrospinal fluid (CSF), 121, 348, 352, 362, 375, 568, 589, 617 Cerebrospinal fluid (CSF) leak, 137, 145, 227, 299, 399, 411, 422, 424 Cerebrovascular complications, 121 Cerebrovascular management
arterial encasement after vascular sacrifice, 208 anesthesia, monitoring, and preparation, 209 bypass grafts choice, 209 subarachnoid encasement, 208 bypass techniques high-flow bypass, 209–211 STA to MCA bypass, 211–212 cavernous sinus meningioma, 218–220 meningioma with subarachnoid vascular encasement, 218 postoperative management, 212 preoperative embolization, 207–208 preoperative imaging, 207 recurrent chondrosarcoma (emergent bypass), 220 staged operations, 212 vasospasm, 217–218 vein graft, 222–223 vein of Labb´e reconstruction, 210f , 220–222 veins and sinuses reconstruction, 213–215 cerebral venous sinuses, 215 intraoperative sinus occlusion test, 215 venous reconstruction, 213t, 215 venous sinuses reconstruction direct repair, 217 graft reconstruction, 217 Cervical lymphadenopathy, 335, 457 Cervical–parotid approach, 337–338, 341f Chemotherapy, 438, 450, 456, 468, 474–477, 491, 508, 566–568, 579, 618 Chemotherapy for skull base tumors for paranasal sinus cancer adenoid cystic carcinoma, 177 esthesioneuroblastoma, 177 neuroendocrine carcinoma, 177–178 sinonasal undifferentiated carcinoma, 176–177 squamous cell carcinoma, 175–176 for sinonasal cancers, 175 Cholesteatomas, 378 Chondroid chordomas, 419, 477 Chondromas, 376 Chondrosarcoma, 54, 87, 168–169, 277–278, 293–294, 345, 367, 368, 375–376 Chondrosarcoma of skull base clinical presentation, 497 diagnosis, 497 epidemiology, 495 imaging, 497 incidence, 495 outcome and prognosis, 500–501 pathology, 496–497 treatment, 497 chemotherapy, 500 radiation, 499–500 surgery, 498–499 Chordoid meningiomas, 505 Chordoma, 53–54, 167–168, 268, 294, 311, 367, 375, 376, 419, 476–477 Chordoma of skull base clinical presentation, 497 diagnosis, 497 epidemiology, 495–496 imaging, 497 incidence, 495–496 outcome and prognosis, 500–501 pathology, 496–497 treatment, 497–500 chemotherapy, 500 surgery, 498–499 radiation, 499–500 Chromosomal abnormalities, cancer, 66–69 Chromosomal staining, 65f Circular sinus, 309 Circumglossal approach, 282–285
Index Clear cell meningioma, 505 Clival chordomas, 71 Clival tumors, 497 clinical assessment, 278–279 diagnostic imaging, pathology, and diagnosis, 277–278 postoperative care and follow-up, 290–291 surgical anatomy, 277 surgical management circumglossal approach, 282–285 extended transbasal approach, 285–287 retropharyngeal approach, 282–285 transcondylar approach, 287–290 transmandibular approach, 282–285 transmaxillary approaches, 280–281 transoral–transpalatal approach, 280 transpalatal approach, 282–285 transpetrous approaches, 287 transsphenoethmoidal approach, 280 transsphenoidal approach, 279–280 symptoms, 279t Clivus, 308–309 Clonidine, 549 Collet–Sicard syndrome, 544–545 Combination chemotherapy, 475 Computed tomography (CT), 81, 82–92, 156, 160, 238, 270, 277, 295, 335, 347, 391, 431, 454, 497, 517–518, 545–546 Computerized dynamic posturography (CDP), 114t Concomitant chemotherapy, 456 Conductive hearing loss, 106 Condylar fossa, 7 Condylectomy, 323 Congenital nystagmus, 97 Constructive interference in steady state (CISS) image, 585f , 590f Conventional chondrosarcoma, 294, 496, 501 Conventional chromosomal analysis, 63 Conventional fractionated radiotherapy, 550 Convexity meningiomas, 507 Corneal reflex, 103 Coronal bone window CT scan, 585f Coronal MRI, 167f Cortisol deficiency, 565 Cranial base anatomy anterior cranial base anterior view, 6f endocranial surface, 3 exocranial surface, 3, 7 oblique anterior view, 5f superior view, 5f middle cranial base, 3, 5 anterior view, 5f , 7f endocranial surface, 7 exocranial surface, 24 oblique anterior view, 5f superior view, 5f , 12f posterior cranial base endocranial surface, 37 exocranial surface, 40 Cranial base reconstruction, 260–261 Cranial nerves. See also under Head, neck, and neuro-otologic assessment of tumor deficits, 545t, 550t dysfunction, 144–145 injury, 183, 184 syndromes, 105t Cranial neuropathies, 551, 617 Craniocervical chordomas, 294 Craniofacial resection, 250–256, 296–297, 455, 456t Cranionasal separation, 260–261 Craniopharyngiom, 359 Craniopharyngioma, 166–167, 357 Craniopharyngioma (neurosurgical management) endocrine dysfunction, 573
hypothalamic dysfunction, 573 intracranial pressure, 573 pathology, 573–574 prognosis, 579 staging, 574–575 treatment adjuvant therapy, 579 hydrocephalus management, 576 neurosurgical approaches, 576–579 preoperative medical management, 576 visual disturbance, 573 Craniotomy, 17f , 324–325, 371 Craniovertebral junction (CVJ) tumors clinical assessment, 420 complications, 424 conclusion, 424 diagnostic studies, 420 pathology and diagnoses, 418–420 surgical anatomy, 417–418 surgical technique retrosigmoid approach, 422–424 transfacial approaches, 422 transoral approach, 420–422 Cruveilhier’s pearly tumors, 583 CSB (central skull base), 82, 84, 87 CSF fistula in temporal bone surgery, 387t CSF. See Cerebrospinal fluid CSF leak. See Cerebrospinal fluid (CSF) leak CT. See Computed tomography Cushing’s disease, 358, 361 Cushing’s syndrome, 561, 562f , 568 Cylindrical cell papilloma, 45, 45f Cystic schwannoma, 515 Cystoperitoneal shunting, 589–590 Cysts of skull base arachnoid cysts classification, 589 clinical manifestations, 589 embryology and pathophysiology, 589 incidence and epidemiology, 588–589 outcome and prognosis, 591 pathology, 589 radiology, 589 treatment, 589–591 dermoids clinical manifestations, 587 embryology, 587 incidence and epidemiology, 586–587 outcome and prognosis, 588 pathology, 587 radiology, 587 treatment, 588 epidermoids clinical manifestations, 584 embryology, 583 incidence and epidemiology, 583 outcome and prognosis, 585–586 pathology, 583–584 radiology, 584 treatment, 585 Cytogenetic terms and abbreviations, 67t Cytotoxic chemotherapeutic agents, 568 DDAVP, 362 Debulking surgery, 438–439 Dedifferentiated chordoma, 70, 72f , 477 Deep-lobe parotid tumors, 339–340 Dental restoration, 264 Dermoids. See under Cysts of skull base Dexamethasone, 127 Diabetes insipidus, 576, 616 Diagnostic chromosomal changes, 69t
621
622
Index
Diffusion-weighted image (DWI), 589 Digital subtraction angiography, 547 Diplopia, 279 Disease-specific instruments, 189–192 DNA-arrays, 66 Dopamine agonist therapy, 567 Doppler ultrasound, 211 Doxorubicin, 474, 475 Dural grafting, 319 Dysarthric speech, 183 Ear. See under Head, neck, and neuro-otologic assessment of tumor EBV (Epstein–Barr virus), 48–49, 294, 316 Ectopic meningiomas, 503 Ectopic pituitary adenoma, 43 Electrocochleography, 123 Electrodiagnostic testing, 106 Electromagnetic radiation, 159 Electromyography (EMG), 122–123 Electron microscopy, 462, 558 Electronystagmography (ENG), 112–114 Emboli in dogs, 306t Embolization, 207–208, 407, 548 Embryonal rhabdomyosarcoma (ERMS), 75 Emergent bypass, 220 ENB. See Esthesioneuroblastoma Encephalocele, 43 Endocrine dysfunction, 573 Endometrial hyperplasia, 508 Endonasal approaches, 244 Endonasal endoscopic skull base surgery complications cerebrospinal fluid leak, 137 infectious complications, 137 tension pneumocephalus, 137 vascular injuries, 136–137 minimal access skull base approaches, 131–134 nasal corridor, 132 panclival approach, 134 transcribriform, 132–133 transpterygoid/infratemporal fossa, 133–134 transsellar/transplanum, 132 nonvascularized reconstruction, 136 principles and philosophy, 131 transpterygoid temporoparietal flap (reconstruction), 135–136 vascularized reconstruction (nasoseptal flap), 135 Endoscopic approach, 126–127, 298–300, 440, 488 Endoscopic endonasal approaches (EEA), 244 Endoscopic resection, 456t Eosinophilic granuloma, 47, 419, 497 Epicenter, 81 Epidermal growth factor receptor (EGFR), 504 Epidermoids. See under Cysts of skull base Epirubicin, 474 Epithelial immunohistochemical markers, 574 Epithelial membrane antigen (EMA), 53, 54, 506, 515 Epithelioid-type melanoma, 460 Epstein–Barr virus (EBV), 48–49, 294, 316 Esophageal, 181 Esthesioneuroblastoma (ENB), 72, 169, 177, 293, 453 incidence and epidemiology, 453 management chemotherapy, 456 clinical features, 454 imaging, 454 radiotherapy, 456 surgery, 455–456 outcome and prognosis, 456–457 pathology, 453–454 staging, 454 Ethmoid carcinoma, 239f Ethmoid malignant tumors, 446t
Ethmoid sinus, 228–229, 233f Ethmoid tumors, 430, 431, 433t Eustachian tube, 306, 326 Ewing’s sarcoma, 55 Exophytic Schneiderian papilloma, 45 Extended transbasal approach, 285–287 External auditory canal (EAC), 345, 370 External beam radiation therapy (EBRT), 507, 551, 565, 579 External carotid artery (ECA), 403, 540 Extracranial approach, 250–251 Extradural clival tumors, 495 Extradural cyst, 587 Extradural zygomatic/middle fossa approach, 520 Extramedullary plasmacytoma, 268 Extraoral lesions, 151 Extraparotid salivary tissue, 333 Eye and orbit. See under Head, neck, and neuro-otologic assessment of tumor Eye movements, 96 Facial defects restoration, 264 Facial nerve, 103, 339f Facial nerve dysfunction, 400 Facial nerve schwannomas (FNS) clinical features 522–523 decision–making, 526–527 diagnosis, 525 management, 525–526 radiation, 534 radiologic characteristics, 523 topographic classification, 525f Facial translocation approaches, 488, 489, 490f Familial paraganglioma, 74, 541 Fast neutron therapy, 467 Fiberoptic endoscopic evaluation of swallowing (FEES), 185 Fibro-osseous lesions bilateral facial neurofibromatosis, 610–613 fibrous dysplasia of craniofacial region, 597–602 neurofibromatosis, 602–610 Wegener granulomatosis, 613–614 Fibrosarcoma, 52, 314, 315 Fibrous dysplasia, 497, 597–602 Fibrous meningiomas, 505 Fine–needle biopsy, 337 Fisch classification, 543t Fisch infratemporal approach, 530 FISH (fluorescent in situ hybridization), 63–65 Flap compromise, 144 Flexible nasal endoscopy, 335 Flexner–Wintersteiner rosette, 55 Fluid-attenuated inversion recovery (FLAIR), 392 Fluorescent in situ hybridization (FISH), 63–65 FNS. See Facial nerve schwannomas Follicle-stimulating hormone (FSH), 557 Foramen magnum syndrome, 22, 527 Foramen ovale exposure, 323 Fractionated radiation therapy, 162t Fractionated stereotactic radiotherapy (FSRT), 160, 164, 165, 566 Frontal craniotomy, 252, 255f Frontal sinus, 229 Frontal–subcortical circuits, 203, 204f Frontonasal approach, 601f Frontotemporal bone flap, 34f Fukuda stepping test, 103 Gait and balance testing. See under Head, neck, and neuro-otologic assessment of tumor Gamma knife plan, 162f , 163, 165 Gamma knife radiosurgery, 295, 579, 618 Gasserian ganglion syndrome, 616 Gaze nystagmus, 97 G-banding technique, 63
Index Gene amplification, 69 Gene therapy, 76 GeneChip arrays, 67f Generic instruments, 189, 190t Genetic abnormalities, 504 Genetic abnormalities of skull base tumors chromosomal abnormalities in cancer, 66–69 clival chordomas, 71 comparative genomic hybridization, 65–66 conventional chromosomal analysis, 63 fluorescent in situ hybridization, 63–65 gene therapy, 76 juvenile nasopharyngeal angiofibroma, 74 meningioma, 72–74 molecular arrays, 66 neurogenic skull base tumors, 70–72 paraganglioma, 74 pleomorphic adenoma, 74 salivary gland tumors, 70–71 sarcomas, 74–76 skull base cytogenetics, 76 spectral karyotyping, 63–65 squamous cell and undifferentiated carcinoma, 69–70 submicroscopic genetic deviations identification, 66 Genomic hybridization (CGH), 65–66 GFAP (glial fibrillary acid protein), 515 GH-secreting adenomas, 358–359 Glasscock–Jackson classification, 543t Glenner and Grimley classification, 540t Glomus jugulare tumor growth complications, 411 clinical presentation, 406 diagnostic imaging, 406, 407 follow-up and rehabilitation, 411 pathology, 404–405 postoperative care, 410 preoperative preparation embolization, 407–408 hormonal studies, 407 intraoperative neurophysiological monitoring, 408 reconstruction, 410–411 surgical approach bone removal, 409 closure, 410 incision and soft-tissue dissection, 408–409 intradural tumor removal, 410 patient position, 408 tumor isolation, 409–410 Glomus tumor, 91, 93f , 167 Glomus tympanicum, 350f , 543t, 549 Glossopharyngeal nerve, 403 Growth hormone (GH), 161, 557 GSPN (greater superficial petrous nerve), 367, 372, 375 Haddad-Bassagasteguy flap (HBF), 135 Head and neck paragangliomas. See Paragangliomas of head and neck Head thrust test, 97 Head, neck, and neuro-otologic assessment of tumor audiometric, vestibular, and electromyographic studies, 103 acoustic neuroma, 107–109, 110–112 audiometry, 106–107 auditory brainstem responses, 109 computerized dynamic posturography, 114–115 electronystagmography, 112–114 otoacoustic emissions, 112 rotatory chair testing, 114 stacked ABR, 112 vestibular evoked potentials, 115 cranial nerves, 100 corneal reflex, 103 facial nerve, 103 olfactory nerve, 100 trigeminal nerve, 100–103
ear pinna and external auditory canal, 97–98 tuning fork examination, 98–99 tympanic membrane and middle ear, 98 eye and orbit eye and orbit, 95 eye movements, 96 head-shaking nystagmus, 97 head thrust test, 97 nystagmus, 96–97 gait and balance testing, 103 Fukuda stepping test, 103 gait, 103 sensory integration and balance, 104 stance, 103 head, scalp, and skin examination, 95 larynx, 99 neck, parotid, and thyroid, 99–100 nose and nasopharynx, 99 oral cavity and oropharynx, 99 Head–shaking nystagmus (HSN), 97 Hearing loss, 106, 107 Hemangiomas and vascular malformations, 483–484 Hemangiopericytoma (HPC), 50–51, 50f , 246f , 313, 477, 481, 483 Hematogenous metastasis, 295 Hematoma, 399 High-dose ifosfamide, 475 High-dose-perfractionation (HDPF) therapy, 466 High-flow bypass, 209–211 Histiocytosis-X (eosinophilic granuloma), 47 Horner syndrome, 617 House–Brackmann facial nerve grading system, 103, 105t, 400 Human papillomavirus (HPV), 45 Hyaline chondrosarcomas, 496 Hydrocephalus, 399, 576 Hydroxyurea, 508 Hypersecretory syndrome, 557, 562 Hypoglossal nerve, 403 Hypoglossal nerve schwannomas, 92f , 532 Hypothalamic dysfunction, 573 Hypothalamic–pituitary–adrenal (HPA), 565 ICA. See Internal carotid artery Ifosfamide, 475 Immittance testing, 107 Immunohistochemical staining, 515, 557 Immunohistochemistry, 506 Immunophenotyping, 52 Induction chemotherapy, 175, 176 Inferior maxillectomy, 248, 252f Inferior turbinectomy, 274f Inflammatory pseudotumor, 43, 44f Infrastructure maxillectomy, 436 Infratemporal fossa, 5f , 6f , 8, 12f , 230, 489, 490f Infratemporal fossa/middle fossa approach, 305 anatomy, 306 cavernous sinus and clivus, 308–309 pterygomaxillary space, 311 sphenoid sinus, 308 temporal bone, 309 combined resection, 325–328 craniotomy, 324–325 foramen ovale exposure, 323 incision, 319–320 intratympanic exposure, 323–324 mandibular condylectomy, 323 pathology, 311–318 surgical procedures lateral approach to skull base, 319 preoperative planning, 319 temporalis muscle dissection, 322 zygomatic ostectomy, 322–323
623
624
Index
Intensity-modulated radiotherapy (IMRT), 160, 438, 499 Internal auditory canal (IAC), 90, 367, 371 Internal carotid artery (ICA), 131, 207, 277, 296, 375, 540 Interstitial brachytherapy, 500 Intra-arterial (IA) chemotherapy, 175, 438 Intrabulbar dissection, 409–410 Intracranial chordomas, 495 Intracranial dermoid cysts, 588 Intracranial extension (ICE), 540 Intracranial infections, 145 Intracranial meningiomas, 81, 503, 508t Intracranial pressure, 573 Intradural cyst, 587 Intradural tumor removal, 410 Intraforaminal schwannoma, 413f Intraoperative neurophysiological monitoring, 408 Intraoperative sinus occlusion test, 215 Intratympanic exposure, 323–324 Invasive adenoma on MR imaging, 564f Invasive fungal sinusitis, 46 Inverted schneiderian papilloma, 45 Jugular foramen, 24 Jugular foramen schwannomas (JFS) diagnosis, 528–529 origin, 527 radiation, 534 radiologic investigation, 527–528 surgical treatment, 530–532 symptoms, 527 treatment options, 529–530 Jugular foramen syndrome, 617 Jugular fossa tumors anatomy, 403 glomus jugulare tumors. See Glomus jugulare tumor growth meningiomas clinical presentation, 413 diagnostic imaging, 413 pathology, 413 preoperative preparation, 413 surgical approach, 414–415 pathology and diagnosis, 403–404 schwannomas clinical presentation, 412 diagnostic imaging, 412 pathology, 412 postoperative course, 413 preoperative preparation, 412 surgical approach, 412–413 Jugular paragangliomas, 540, 549–550 Jugular schwannoma, 92f Jugulotympanic paragangliomas, 543t, 544–545 Juvenile angiofibroma, 84, 86f , 267, 314f Juvenile nasopharyngeal angiofibroma (JNA), 74, 243f . See also under Vascular tumors of skull base Kadish staging system, 177t, 294t Kaplan Meier survival, 457f Klebsiella rhinoscleromatis, 46 Lacrimal system, 232, 236f Laryngeal paragangliomas, 540, 545 Laryngoplasty, 342 Larynx, 99 Lateral position for surgery, 120 Lateral rhinotomy, 245, 248, 249f Laterocerebellar cysts, 589 Leiomyosarcoma, 313, 315f Leksell gamma knife system, 533, 534 Lesser superficial petrous nerve (LSPN), 372 Liposarcoma, 476
Lobular capillary hemangioma, 49–50 L-shaped craniotomy, 325 Luteinizing hormone (LH), 557 Lymphatics, 227 Lymphoepithelioma, 316 Lymphoma, 57, 296–297 Lymphoproliferative disorders. See under Tumor and tumor-like lesions of skull base Macroadenoma, 557 Macroglossia/facial swelling, 121 Maffucci syndrome, 496 Magnetic resonance angiography (MRA), 546 Magnetic resonance imaging (MRI), 81–92, 160, 167f , 168f , 215f , 218f , 238, 270, 277, 294, 295, 355, 359, 368f , 391–393, 454, 497, 517, 518, 546–547, 617 Magnetic resonance venography (MRV), 546 Malignant epithelial neoplasms. See under Tumor and tumor-like lesions of skull base Malignant fibrous histiocytoma (MFH), 65 Malignant melanoma, 294 Malignant meningiomas, 505–506 Malignant neoplasms. See under Tumor and tumor-like lesions of skull base Malignant paragangliomas, 541–542 Malignant peripheral nerve sheath tumor (MPNST), 72, 517 Malignant schwannoma, 517 Malignant sinonasal tumors, 44t Malignant tumors, 268 Mandibular condylectomy, 323 Mandibular fossa, 8f Mandibulotomy, 283, 340f , 341 Mastoidectomy, 36f , 38f , 290, 349, 350 Maxilla, 8f , 33f , 150–151 Maxillary reconstruction, 151 Maxillary sinus, 228, 233f , 433t Maxillofacial prosthodontist, 240 Maxillotomy, 34 McCabe/Fletcher classification, 544t McCune–Albright syndrome, 597, 600, 602 Meckel cave, 5, 85, 220, 325, 369, 519 Medial maxillectomy, 248, 249f , 434, 435f Melanoma, 56–57 Melanoma of nasal cavity and paranasal sinuses clinical presentation and findings, 463 epidemiology, 459–460 malignant melanoma, 462t, 464t outcome and prognosis, 468–469 pathology, 460–463 staging, 463–465 treatment, 465 chemotherapy and immunotherapy, 468 radiation, 466–468 surgery, 465–466 Meningioma, 58, 72–74, 164–166, 218, 293, 345, 368–369, 378, 418, 497, 503 adjuvant therapy, 508–509 asymptomatic meningioma, 508 chemotherapy, 508 etiology infection, 503 radiation, 503 trauma, 503 genetics genetic abnormalities, 504 NF-2 mutations, 503–504 immunohistochemistry, 506 incidence, 503 pathology, 504–505 radiation external beam radiation therapy, 507 stereotactic radiosurgery and radiotherapy, 507–508 with subarachnoid vascular encasement, 220
Index surgery classification, 505t convexity meningiomas, 507 olfactory groove, 507 parasagittal meningiomas, 507 posterior fossa meningiomas, 507 sphenoid wing meningiomas, 507 tuberculum sellae meningiomas, 507 tumor biology growth factors, 504 receptors, 504 WHO Grade I, 505 WHO Grade II, 505 WHO Grade III, 505–506 Meningitis, 399–400 Meningo-encephalocele, 245f Meningothelial meningioma, 504, 505 Mesenchymal tumors. See under Tumor and tumor-like lesions of skull base Mesoadenoma, 557 Metastasis, 378 Metastatic neoplasms, 59 Metastatic skull base tumors diagnosis of skull base metastases, 617 jugular foramen syndrome, 617 middle fossa syndrome, 616–617 occipital condyle syndrome, 617 orbital syndrome, 615 prognosis, 617 sellar/parasellar syndrome, 615–616 temporal bone involvement, 617 treatment, 617–618 Microadenomas, 557 Microprolactinomas, 564 Microvascular reconstruction. See Prosthetic rehabilitation (skull base injury) Middle cranial base. See under Cranial base anatomy Middle cranial fossa resection, 250 Middle cranial fossa tumors complication avoidance, 373–374 clinical assessment, 369 diagnostic imaging, 369–370 follow-up and rehabilitation, 374 pathology and diagnosis chondrosarcoma, 368 chordoma, 367 meningioma, 368–369 trigeminal schwannoma, 369 postoperative care, 373 preoperative preparation, 370 surgical anatomy, 367 surgical technique approach selection, 370 closing, 373 dural opening, 372–373 execution, 371 middle fossa exposure, 371 petrous apex drilling, 372 positioning, 370–371 surface landmarks, 370 tumor resection, 373 Middle fossa approach, 35f , 381–383 Middle fossa syndrome, 616–617 Middle fossa/infratemporal fossa resection (combined), 325–328 Middle skull base, 142, 183 Midfacial degloving (MFD), 247, 297, 435f Minimally invasive techniques. See Endonasal endoscopic skull base surgery Minimum alveolar concentration (MAC), 123, 124 Mixed hearing loss, 106 Modified barium swallow (MBS), 185, 186 Molecular arrays, 66
625
Motor control test (MCT), 114 Motor root dysfunction, 616 MRI. See Magnetic resonance imaging Mucosal melanoma, 57t Multidisciplinary team approach, 197 Myocutaneous flaps, 299 Myocutaneous, 141 Myospherulosis, 47 Myxoid chondrosarcomas, 496 Myxoma, 51 Nasal cavity, 3, 5 Nasal cavity tumors anatomy lymphatics, 225 nasal cavity arteries, 227 nasal cavity floor, 227 nasal cavity roof, 227 nasal wall, 227 sensory nerves, 227 venous drainage, 227 biopsy, 239–240 clinical assessment, 236–237 imaging, 237–239 pathology and diagnosis, 232–236 preoperative preparation, 240–242 reconstruction and rehabilitation cranial base reconstruction (cranionasal separation), 260–261 dental restoration, 264 facial defects restoration, 264 orbito-maxillary reconstruction (eye and cheek support), 261–264 palatal reconstruction (oronasal separation), 258–260 resection craniofacial resection, 250–256 inferior maxillectomy, 248 medial maxillectomy, 248 total maxillectomy, 248–250 surgical approach endonasal approaches, 244 midfacial degloving, 247 sublabial approaches, 247–248 transfacial approaches, 245 surgical principles, 242–244 Nasal corridor, 132 Nasal glial heterotopia, 43 Nasal mucosal melanoma. See Melanoma of nasal cavity and paranasal sinuses Nasoethmoid chondrosarcoma, 90f Nasopharyngeal angiofibroma, 51, 51f , 314f Nasopharyngeal carcinoma (NPC), 48–49, 84, 87f , 169, 268, 269, 278, 294, 306, 316 Nasopharyngeal malignancies, 316 Nasopharynx tumors clinical assessment, 268–270 computed tomography, 270 magnetic resonance imaging, 270 PET/CT scan, 270 serological tests, 270 pathology and diagnosis benign tumors, 267 malignant tumors, 268 surgical anatomy, 267 surgical resection, 270–271 surgical technique bony dissection and osteotomies, 272–273 closure, 274–275 complications and avoidance, 275 follow–up and rehabilitation, 275 incision, 272 patient position, 272 postoperative care, 275
626
Index
Nasopharynx tumors (Continued) soft tissue dissection, 272 tumor resection, 273 Nasoseptal flap, 135 Neck management, 440 Neck nodes classification, 433t Nelson syndrome, 161, 568 Neoplasms of parapharyngeal space, 333t Neural (Retrocochlear) hearing loss, 108t Neurilemmomas, 333, 334 Neuroblastoma, 55 Neurocognitive assessment ABST patients, 201 frontal–subcortical circuits, 203 neuropsychological assessment, 203–205 pharmacologic and behavioral interventions, 205 posttreatment studies, 202–203 pretreatment studies, 201 treatment effects, 201–202 tumor location, 201 Neurocutaneous disorders, 96t Neuroendocrine carcinoma, 177–178 Neuroendocrine neoplasms. See under Tumor and tumor-like lesions of skull base Neurofibroma, 71, 315, 334, 514–515 Neurofibromatosis, 602–610 Neurogenic neoplasms. See under Tumor and tumor-like lesions of skull base Neurogenic tumors, 333–334 Neurogenic skull base tumors, 71–72 Neuro-ophthalmologic testing, 573 Neurosurgical approaches for craniopharyngioma, 576–579 NF-2 mutations, 503–504 Non-diagnostic chromosomal changes, 70t Non-Hodgkin lymphomas, 294 Nonmelanotic melanoma, 465f Nonmelanotic nasopharyngeal mucosal melanoma, 461f Nonsalivary gland adenocarcinoma, 49 Nonsquamous cell carcinoma of nasal cavity and paranasal sinuses adenocarcinoma clinical features, 449 outcome and prognosis, 450–451 pathology, incidence, and epidemiology, 448–449 treatment, 449–450 adenoid cystic carcinoma clinical features, 447 incidence and epidemiology, 446–447 outcome and prognosis, 447–448 pathology, 447 treatment, 447 clinical features, 445 incidence and epidemiology, 445 pathology, 445 staging, 445–446 Nonsteroidal anti-inflammatory drugs (NSAIDs), 197 Nonsyndromic familial paragangliomas, 541 Nonvascularized reconstruction, 136 Nonvestibular schwannomas (NVS) anatomic distribution, 518 history, 518–519 in children, 519 radiation, 533 symptoms, 519 Nose and nasopharynx, 99 Null-cell adenomas, 558 NVS. See Nonvestibular schwannomas Nystagmus, 96–97 Obturator interim, 259, 263f permanent, 259, 263f surgical, 259 Occipital bone, 24
Occipital condyle syndrome, 617 Occipito-transmastoid-cervical approach, 384–386 Octreotide scintigraphy, 547 Ocular motility nerves schwannomas, 532 Ohngren line, 430, 431f , 446 Olfactory and optic nerve schwannomas, 532 Olfactory esthesioneuroblastoma, 462 Olfactory groove, 507 Olfactory groove meningioma (OGM), 73, 82f Olfactory nerve, 100, 228, 253 Olfactory neuroblastoma, 83f , 453, 455f , 457f Oncogenes, 68 Oncomitant chemoradiotherapy, 176, 178 Oral and dental anatomy, 149 Oral cavity and oropharynx, 99 Oral preparatory, 181 Orbit tumor anatomy, 231 biopsy, 239–240 clinical assessment, 236–237 imaging, 237–239 orbit management, 439 primary orbital tumors, 257–258 sinonasal tumors, 256–257 pathology and diagnosis, 232–233 preoperative preparation, 240–242 reconstruction and rehabilitation cranial base reconstruction (cranionasal separation), 260–261 dental restoration, 264 facial defects restoration, 264 orbito-maxillary reconstruction (eye and cheek support), 261–264 palatal reconstruction (oronasal separation), 258–260 resection craniofacial resection, 250–256 inferior maxillectomy, 248 medial maxillectomy, 248 total maxillectomy, 248–250 surgery, 242–244 endonasal approaches, 244 midfacial degloving, 247 sublabial approaches, 247–248 transfacial approaches, 244–245 Orbital osteotomies, 252 Orbital syndrome, 615 Orbitofrontal craniotomy, 29f Orbito-maxillary reconstruction, 261–264 Orbitozygomatic approach, 297 Orbitozygomatic-temporal approach, 521f Oronasal separation, 258–260 Oropharynx, 99 Osseointegrated dental implants, 153–156 Osseous cross section, 8 Osteoblastomas, 497 Osteomas, 497 Osteosarcoma, 54, 55f , 76, 476 Osteotomies, 29f , 248, 249f , 349 Otoacoustic emissions (OAE), 112 Palatal reconstruction, 258–260 Palliative chemoradiotherapy, 348 Panclival approach, 134 Papillary meningioma, 506 Paraganglioma, 56, 74, 334, 350t, 406f , 407, 479, 528 Paraganglioma nomenclature, 539 Paragangliomas of head and neck carotid body tumors, 543 jugulotympanic paragangliomas, 544 laryngeal and sinonasal paragangliomas, 545 vagal paragangliomas, 545 classification and staging, 543 diagnosis, 543
Index diagnostic imaging computed tomography, 545–546 diagnostic angiography, 547 magnetic resonance imaging, 546–547 octreotide scintigraphy, 547 ultrasound, 545 epidemiology and etiology genetic counseling, 541 genetics of paragangliomas, 541 location and routes of spread, 539–540 malignant paragangliomas, 541–542 management, 547–548 natural history, 542–543 paraganglioma nomenclature, 539 paraganglion system, 539 pathology, 541 physiology and function, 539 radiation, 551 complications, 551–552 conventional fractionated radiotherapy, 550–551 stereotactic radiosurgery, 551 secreting paragangliomas, 542 surgery bilateral carotid body tumors, 548–549 for carotid body paragangliomas, 548 for glomus tympanicum, 549 for jugular paragangliomas, 549–550 for vagal paragangliomas, 549 preoperative embolization, 548 superselective angiography, 548 Paranasal mucosal melanoma. See Melanoma of nasal cavity and paranasal sinuses Paranasal sinus cancer treatment. See under Chemotherapy for skull base tumors Paranasal sinus tumors, 169 anatomy, 228–230 clinical assessment, 236–237 imaging, 237–239 pathology and diagnosis, 232–233 preoperative preparation, 240–242 reconstruction and rehabilitation cranial base reconstruction, 260–261 dental restoration, 264 facial defects restoration, 264 orbito-maxillary reconstruction, 261–264, 264f palatal reconstruction, 258–260 resection craniofacial resection, 250–256 inferior maxillectomy, 248 medial maxillectomy, 248 total maxillectomy, 248–250 surgical approach endonasal approaches, 244 midfacial degloving, 247 sublabial approaches, 247–248 transfacial approaches, 244–245 surgical principles, 242–243 Parapharyngeal space tumors clinical assessment, 334–335 complications and avoidance, 342 diagnostic imaging, 335–337 follow-up and rehabilitation, 342 pathology and diagnosis, 332–334 miscellaneous tumors, 334 neurogenic tumors, 333–334 salivary gland neoplasms, 333 postoperative care, 341–342 preoperative preparation, 337 surgical anatomy, 331–332 surgical technique bony dissection and osteotomies, 339 closure, 341 incision, 338
positioning, 338 reconstruction, 341 soft tissue dissection, 338 tumor resection, 339–341 Parasagittal meningiomas, 507 Parasellar anatomy, 355 Park-bench position (lateral oblique position), 119 Parotid salivary duct carcinoma, 89f Particle beam therapy, 159 Patient controlled analgesia (PCA), 127 Pediatric sarcomas, 476 Pedicled myocutaneous flap, 139 Pegvisomant, 567 Pericranial incisions, 252 perineural tumor spread (PNS), 84 Peripheral nerve/cranial nerve injuries, 122 Peroxisome proliferator-activating receptor (PPAR), 568 PET scan, 295–296 PET-CT images, 242f , 270 Petroclival approach, 131 Petroclival meningioma, 378f Petro-occipital trans-sigmoid (POTS) approach, 530–531 Petrous apex approach, 131 Petrous apex erosion, 519 Petrous apex lesions on CT scanning, 379t on MRI scanning, 380t Petrous apex tumors, 375 clinical assessment, 378 complications and avoidance, 387–388 diagnostic imaging, 378–380 pathology and diagnosis cholesteatomas, 378 chondromas, 376 chondrosarcomas, 375–376 chordomas, 376 meningiomas, 378 metastasis, 378 schwannomas, 378 postoperative care, 386–387 preoperative preparation, 380 surgical anatomy, 375 surgical approaches middle fossa approach, 381–383 occipito-transmastoid-cervical approach, 384–386 posterior fossa (retrosigmoid) approach, 383 presigmoid approach, 383–384 temporal bone resection, 386 translabyrinthine approach, 380–381 Petrous Apex, drilling of, 372 Petrosectomy, 287 Pharyngeal flap, 284 Pharyngeal swallowing, 183, 185 Phosphorylated histone H3 (PHH3), 506 Physaliphorous cell, 313 Pinna and external auditory canal, 97–98 Pituitary adenomas, 44f , 86, 161–163 chemotherapy—medical therapy, 566–568 clinically nonfunctioning adenomas, 358 Cushing disease, 358 GH-secreting adenomas, 358–359 by hormonal production, 558t incidence and epidemiology, 557 pathology, 557–560 PRL-secreting adenomas, 358 radiation, 565–566 staging, 560–561 surgery, 563–565 treatment, 561 history and physical examination, 561 laboratory tests, 562 radiology, 562–563 by tumor type, 558t
627
628
Index
Pituitary apoplexy on MRI, 11f Pituitary macroadenomas, 278, 564f Pituitary tumor transforming gene (PTTG), 560 Plasmacytomas, 419, 497 Platelet-derived growth factor (PDGF) receptor, 504 Pleomorphic adenoma, 74 Pleomorphism, 419 Pneumocephalus, 121 Polymerase chain reaction (PCR), 65 Polyomavirus, 503 Positional nystagmus, 97 Positive end-expiratory pressure (PEEP), 121 Positron emission tomography (PET), 239, 463, 617 Posterior cranial fossa, 142 Posterior fossa (retrosigmoid) approach, 383 Posterior fossa meningiomas, 507 Posterior Inferior Cerebellar Arteries (PICA), 40, 540 Posterior septectomy, 132 Posterior skull base, 183, 184 Posteroinferior cerebellar artery (PICA), 40, 410 Postoperative physiotherapy, 150 Postoperative radiotherapy, 500, 501 Postradiation bypass, 218–220 Poststyloid compartments, 333t Prechiasmatic lesions, 577 Premaxilla, 150 Presigmoid approach, 38f , 383 Prestyloid compartments, 333t Primary clival chordoma, 71, 72 Primary orbital tumors, 257–258 Primary petrous apex lesions, 376t Primitive neuroectodermal tumor (PNET), 48, 55, 56f Progesterone receptors, 504 Prolactin (PRL), 557 Prolactinomas, 566–567 Prone position, 119 Prosthetic rehabilitation (skull base injury) maxilla, 150–151 microvascular reconstruction techniques ear, 152–153 extraoral lesions, 151 nose, 152 orbit, 152 soft palate, 151 oral and dental anatomy, 149 oral and dental evaluation, 149–150 osseointegrated dental implants, 153–156 preprosthetic principles, 150 Protein-arrays, 66 Proton beam irradiation, 499 Proton beam radiotherapy, 166f Psychotherapy and antidepressants, 197 Pterygoid process, 14f Pterygomaxillary space, 311, 313f Pterygopalatine fossa (PPF), 5f , 6f , 8f , 14f , 16f , 230–231, 314f , 465 Pure tone audiogram (PTA), 106, 390 Pyogenic granuloma, 50f Quality of life (QOL) (of tumor patients) ASBS QOL questionnaire, 195t estimation methods disease-specific instruments, 189–192 generic instruments, 189, 190t functional disabilities, 191 improving QOL multidisciplinary team approach, 197 pain control, 195–197 psychotherapy, antidepressants and group support, 197 skull base tumor patients with acoustic schwannoma, 194 with anterior skull base tumors, 192–194
quality of life by proxy, 194–195 surgery on QOL measures, 193f Radiation on cognitive functioning, 202t Radiation therapy, 150, 579, 618 Radiation therapy for skull base tumors advanced radiation therapy techniques, 160–161 clinical applications acoustic schwannoma, 163–164 chondrosarcoma, 168–169 chordoma, 167–168 craniopharyngioma, 166–167 glomus tumor, 167 meningioma, 164–166 nasopharyngeal carcinoma, 169 paranasal sinus tumors, 169 pituitary adenomas, 161–163 normal tissue tolerance, 160 radiation toxicity, 169 radiobiology, 159–160 Radiation toxicity, 169 Radiobiology of skull base tumors. See Radiation therapy for skull base tumors Radiology for arachnoid cysts, 589 for dermoids, 587, 587f , 588f , 589f for epidermoids, 584, 584f , 585f Radionuclide bone scans, 617 Radiosurgery, 418–419 Radiotherapy, 437–438, 447, 456, 466–468, 473–474, 489–491, 507–508, 550 Radkowski modification of JNA, 484t Rectus abdominis, 142, 299 Recurrent chondrosarcoma, 220 Recurrent clival chordoma, 72f Respiratory epithelial adenomatoid hamartoma, 43, 44f Retinoblastoma, 235 Retinoic acid receptor beta (RARB), 579 Retinoic acid receptor gamma (RARG), 579 Retrochiasmatic craniopharyngioma, 577f Retrojugular approach, 414, 415f Retrolabyrinthine approach, 36f Retropharyngeal approach, 282–285 Retrosigmoid approach, 383, 422–424 bony dissection, 395 closure, 396 incision and soft tissue dissection, 394 positioning, 394 tumor resection, 395–396 Revascularization for tumors, 209t Reverse transcriptase–polymerase chain reaction (RT-PCR), 76 Rhabdomyosarcoma (RMS), 51, 52t, 235, 278, 313 Rhabdomyosarcoma and Ewing’s sarcoma, 476 Rhinoscleroma, 46–47 Right hemisphere disorder (RHD), 184 Rinne test, 99 Robotic assisted endoscopic surgery, 246f Rosai–Dorfman disease, 57 Rosenmuller fossa, 16f Rotatory chair testing, 114 Sagittal images, 238, 355 Sagittal postcontrast MRI imaging, 279f , 281f Sagittal T1-weighted MRI, 588f , 591f Salivary gland neoplasms, 333 Salivary gland tumors, 70–71 Salivary-type adenocarcinomas, 49 Sarcomas, 74–76 Sarcomas of skull base chemotherapy anthracyclines, 474–475 chordoma, 476–477
Index combination chemotherapy, 475 hemangiopericytoma, 477 high-dose ifosfamide, 475 ifosfamide, 475 liposarcoma, 476 osteosarcoma, 476 pediatric sarcomas, 476 second-line chemotherapy, 475–476 synovial sarcoma, 476 diagnosis, 473 epidemiology, 473 radiation treatment, 473–474 surgery, 473 treatment at specialty centers, 473 SCC. See Squamous cell carcinoma SCC of nasal cavity and paranasal sinuses epidemiology, 429 etiology, 429–430 outcomes and prognosis, 440–441 pathology and natural history, 430–431 prognostic factors, 433 staging, 433 symptoms and signs, 431 treatment, 433 chemotherapy, 438 debulking surgery, 438–439 endoscopic surgery, 440 neck management, 440 orbit management, 439–440 radiotherapy, 437–438 skin or cartilage invasion management, 440 surgery, 434–437 tissue diagnosis, 432 workup and assessment imaging, 431–432 multidisciplinary consultation, 432 Schneiderian membrane, 232 Schneiderian papillomas, 45 Schwannoma, 315, 316f , 333, 345, 378, 412–413, 514 Schwannomas of skull base clinical features facial nerve schwannomas, 522–527 hypoglossal nerve schwannomas, 532 jugular foramen schwannomas, 527–532 nonvestibular schwannomas, 518–519 ocular motility nerves schwannomas, 532 olfactory and optic nerve schwannomas, 532 specific anatomic sites, 519 trigeminal nerve schwannomas, 519–522 genetics and molecular features, 515–516 immunohistochemistry and special stains, 515 malignancy, 517 management, 535 neurofibromas, 514–515 neurofibromatosis, 513–514 radiation in treatment, 532–534 for facial schwannomas, 534 for jugular foramen schwannomas, 534 for nonvestibular schwannoma, 533 SRS for schwannomas, 534 for trigeminal nerve schwannomas, 533–534 radiographic features CT, 517–518 MRI, 517, 518 schwannomas, 514 schwannomatosis, 514 Secondary petrous apex lesions, 376t Second-line chemotherapy, 475–476 Seldinger technique, 135–136 Sella turcica lesions, 561t Sellar lesion on imaging, 576t Sellar tumors. See also Pituitary adenomas
629
clinical presentation, 357–358 complications and avoidance, 362 diagnostic imaging, 355–356 epidemiology, 356–357 extended transsphenoidal approaches, 361 follow-up, 362 medical and surgical treatment, 358–359 postoperative care, 362 sellar and nasal reconstruction, 362 surgical anatomy, 355–356 transsphenoidal surgery, 359–361 Sellar/parasellar syndrome, 615–616 Sellotomy, 132 Sensorineural anosmia, 100 Sensorineural hearing loss, 106, 108 Serological tests, 270 Shamblin classification for CBT, 543f Sigmoid sinus, 4, 222, 403 Single photon emission computed tomography (SPECT), 277, 617 Sinonasal adenocarcinoma, 49, 51t Sinonasal cancers, chemotherapy for, 175 Sinonasal malignancies, 81 Sinonasal melanomas, 241f , 459, 460t, 464. See also Melanoma of nasal cavity and paranasal sinuses Sinonasal paragangliomas, 545 Sinonasal polyps, 44–45 Sinonasal tract tumors, 237t Sinonasal tumors, 256–257 Sinonasal undifferentiated carcinoma (SNUC), 48, 69, 169, 176–177 Sinus histiocytosis, 57 Sitting position for surgery, 120 Skin or cartilage invasion management, 440 Skull base anatomy, 306–311 Skull base cytogenetics, 76 Skull base defects reconstruction algorithm, 140f anatomy, 139–140 complications cerebrospinal fluid leak, 145 cranial nerve dysfunction, 144–145 flap compromise, 144 intracranial infections, 145 wound infection, 144 history, 139 reconstruction skull base zone I, 140–142 skull base zone II, 142 skull base zone III, 142–143 technical considerations, 143–144 Skull base metastases, 616t, 617 Skull base neoplasms imaging anterior cranial base, 81–82 central skull base, 82–87 imaging modality, 81 posterior cranial base, 87–93 Small cell carcinoma, 56, 56f Soft palate, 151 Soft tissue dissection, 74, 272, 338, 349 Somato-autonomic nerves, 227 Somatosensory evoked potential (SSEP), 123, 124–125 Somatostatin analogs, 567 Spectral karyotyping (SKY), 63–65, 69, 71, 75 Speech and swallowing impairments case studies, 186 communication disorders anterior skull base, 183–184 middle/posterior skull base, 184 communication examination, 185 neurophysiology communication, 182 swallowing, 181–182 swallowing assessment
630
Index
Speech and swallowing impairments (Continued) fiberoptic endoscopic evaluation, 185 modified barium swallow study, 185 swallowing disorders anterior skull base, 182–183 middle skull base, 183 posterior skull base, 183 treatment, 185–186 Speech audiometry, 107 Speech discrimination score (SDS), 107, 108, 109, 391 Speech impairments, 182t, 184, 186 Speech reception threshold (SRT), 107, 162, 508 SPGR technique, 356 Sphenoid sinus, 3, 5, 229–230, 308, 355 Sphenoid wing meningiomas, 503, 507 Sphenoidotomy, 132, 134, 135, 359–360 SPS (superior petrous sinus), 372–373 Squamous carcinoma, 48f Squamous cell carcinoma (SCC), 66, 69, 88f , 175–176, 317f , 345, 351t Standard retrosigmoid craniotomy, 507 Stereolithography (SL), 156 Stereotactic radiosurgery (SRS), 160, 500, 507–508, 509, 534, 551, 566 Stereotactic radiotherapy (SRT), 507–508 Sternocleidomastoid (SCM), 100, 383–385 Stripping surgery, 526 Sturge-Weber syndrome, 484 Styloid diaphragm, 14f Stylomandibular fascia, 339f Stylomandibular ligament, 331, 333 Subarachnoid encasement, 208 Subcranial approach, 297 Subdiaphragmatic lesions, 574 Subfrontal approaches, 252, 255f , 256f Subglottic paragangliomas, 540 Sublabial approaches, 247–248, 250f , 251f Suboccipital approach, 530–531 Suboccipital craniectomy, 27f Subplatysmal flaps, 283 SuccinatedehydrogenaseD (SDHD), 74, 541 Superior cerebellar artery (SCA), 40 Superior orbital fissure (SOF), 231, 252, 257 Superior petrosal sinus (SPS), 371, 372–373 Superselective angiography, 548 Supine position, 119, 272, 285, 288 Supradiaphragmatic lesions, 574 Supraglottic paragangliomas, 545 Suprajugular approach, 412, 413f , 414 Suprameatal triangle, 36f Suprastructure maxillectomy, 436, 436f Swallowing, 181–182, 185. See also Speech and swallowing impairments Synovial sarcoma, 52, 75, 476 Temporal bone, 22f , 309–311, 617 Temporal bone neoplasms, 346t Temporal bone paragangliomas, 543t, 544t Temporal bone resection, 350f , 351, 351f , 386 Temporal bone surgery, 349f Temporal bone tumors, 89 clinical assessment, 345–347 complications and avoidance, 352 diagnostic imaging, 347 follow-up and rehabilitation, 352 postoperative care, 352 preoperative preparation and assessment, 347–348 regional pathology and diagnosis, 345 surgical anatomy, 345 surgical technique bony dissection and osteotomies, 349–352 patient positioning, 348 principles, 348
soft tissue dissection, 349 surgical incision, 348–349 Temporal lobe necrosis (TLN), 203 Temporalis muscle dissection, 322 Tension pneumocephalus, 137 Teratocarcinosarcoma, 58, 58f Therapeutic imaging, 347 Thyroid hormone, 565 Thyroid paragangliomas, 540 Thyroid–stimulating hormone (TSH), 359, 557 Time of flight (TOF), 546 Tissue diagnosis biopsy, 432 histology, 432 Tissue microarray immunohistochemistry (TMA–IHC), 504, 506 Tissue tolerance, 160 TNS. See Trigeminal nerve schwannomas Total maxillectomy, 248–250, 436–437, 437f Trabectidin, 476 Tracheostomy, 339, 342, 420 Transcochlear approaches. 36f Transcondylar approach, 24f , 27f , 287–290 Transcortical approaches, 578 Transcranial approach, 563 Transcranial–transsphenoidal approach, 298 Transcribriform, 132–133 Transesophageal echocardiography (TEE), 120, 121 Transfacial approaches, 244–247, 422 lateral rhinotomy, 245, 248, 249f Weber-Fergusson, 245, 248, 250 Transient otoacoustic emissions (TOAE), 112 Transitional meningioma, 505 Transjugular approach, 414–415 Translabyrinthine approach, 36f , 380–381, 396–399 bony dissection, 397–398 closure, 399 incision and soft tissue dissection, 396 positioning, 396 tumor resection, 398–399 Translocations, 67–69, 75 Transmandibular approach, 282–285 Transmaxillary approaches, 33f , 280–281 Transoral approach, 420–422 Transoral–transpalatal approach, 280 Transpalatal approach, 282–285 Transpetrous approaches, 287 Transpterygoid temporoparietal flap, 135–136 Transpterygoid/infratemporal fossa, 133–134 Transsellar/transplanum, 132 Transsphenoethmoidal approach, 280 Transsphenoidal approach, 279–280, 563 Transsphenoidal resection, 358, 362 Trans-sphenoidal technique, 359–361, 578 Transverse facial incision, 272 Trigeminal nerve, 100–103, 584 Trigeminal nerve schwannomas (TNS), 379, 519–520 management, 520 radiation, 533–534 radiographic appearance, 520 staging, 520 surgical treatment, 520–522 symptoms, 519 Trigeminal schwannoma, 369, 520 Tuberculum sellae meningiomas, 507 Tumor and tumor-like lesions of skull base ameloblastoma, 58 biopsies and frozen sections, 43 inflammatory and granulomatous conditions allergic fungal sinusitis, 45–46 histiocytosis-X (eosinophilic granuloma), 47 invasive fungal sinusitis, 46 myospherulosis, 47
Index Tumor and tumor (Continued) rhinoscleroma, 46–47 Wegener granulomatosis, 47 lymphoproliferative disorders lymphoma, 57 sinus histiocytosis, 57 malignant epithelial neoplasms nasopharyngeal carcinoma, 48–49 nonsalivary gland adenocarcinoma, 49 salivary-type adenocarcinomas, 49 sinonasal adenocarcinoma, 49 sinonasal undifferentiated carcinoma, 48 squamous carcinoma, 48 melanoma, 56–57 meningioma, 58 mesenchymal tumors angiosarcoma, 53 benign tumors, 49 chondrosarcoma, 54 chordoma, 53–54 fibrosarcoma, 52 hemangiopericytoma, 50–51, 50f lobular capillary hemangioma, 49–50 myxoma, 50 nasopharyngeal angiofibroma, 50 osteosarcoma, 54 rhabdomyosarcoma, 51 synovial sarcoma, 52, 52f metastatic neoplasms, 59 neuroendocrine neoplasms, 55–57 carcinoid tumors, 56 paraganglioma, 56 small cell carcinoma, 56 neurogenic neoplasms Ewing sarcoma, 55 neuroblastoma, 55 primitive neuroectodermal tumor, 55 non-neoplastic and congenital lesions cylindrical cell papilloma, 45, 45f ectopic pituitary adenoma, 43 encephalocele, 43 exophytic schneiderian papilloma, 45 inflammatory pseudotumor, 43 inverted schneiderian papilloma, 45 nasal glial heterotopia, 43 respiratory epithelial adenomatoid hamartoma, 43 schneiderian papillomas, 45 sinonasal polyps, 44–45 teratocarcinosarcoma, 58 Tumor isolation, 409–410 Tumor resection, 273, 290, 298–299, 339–341 Tumor suppressor gene, 67 Tumors of parapharyngeal space. See Parapharyngeal space tumors Tuning fork examination, 98–99 Tympanic cavity, 32 Tympanic membrane and middle ear, 98 Tympanic paragangliomas, 350, 540, 544–545 Tympanomastoidectomy, 348
Weber test, 99 Weber–Ferguson incision, 151, 297, 434f , 435f , 436 Weber-Fergusson, 245, 248, 250, 272 Wegener granulomatosis, 47, 613–614 Wound infection, 144
Ultrasound, 337, 545 Unilateral orbital osteotomy, 577f Unilateral subfrontal approach, 577f
Zellballen, 539 Zygomatic ostectomy, 319, 322–323 approach, 319
Vagal paraganglia, 539, 540, 545, 549, 550 Vagus nerve, 403, 539 Vascular encasement, 296 Vascular endothelial growth factor–A (VEGF-A), 504 Vascular injuries, 136–137 Vascular lesions classification, 483t Vascular sacrifice, 208 Vascular tumors of skull base angiography, 485–486 chemotherapy, 491 epidemiology, 481 hemangiomas and vascular malformations, 483–484 hemangiopericytoma, 483 incidence, 481 juvenile nasopharyngeal angiofibroma (JNA) anatomy, 482–483 etiology and pathogenesis, 481–482 pathology, 483 physical examination, 485 prognosis, 491 radiology, 485 radiation therapy, 489–491 staging, 484 surgery endoscopic approach, 488 facial translocation approach, 489 infratemporal fossa approach, 489 medical maxillectomy, 488–489 transpalatal approach, 488 symptoms, 484–485 treatment, 487–488 Vascularized reconstruction, 136 Vasospasm, 121, 217 Vein of Labb´e reconstruction, 210f , 220–222 Venous air embolus (VAE), 120, 120t Venous drainage, 227 Venous reconstruction, 213t, 215 Venous sinuses reconstruction. See under Cerebrovascular management Vernet syndrome, 527 Verocay body, 315, 515 Vertigo, 400 Vestibular evoked myogenic potentials (VEMP), 115 Vestibular schwannoma, 89, 92f Vestibular-ocular reflex (VOR), 97 Viral oncolytic therapy, 76 Virchow’s theory, 583 Visual disturbance, 573 Volatile anesthetics, 125, 126 von Recklinghausen’s disease, 315, 334, 515 Von Remak’s theory, 583
631
about the book… The management of tumors in and adjacent to the skullbase is challenging given the complex and critically important anatomy of the region and the wide diversity of tumor pathologies that may be encountered. To help navigate the complexities of contemporary multidisciplinary management of these patients, Drs. Hanna and DeMonte bring you Comprehensive Management of Skull Base Tumors, a comprehensive guide filled with updated information from authorities around the world. Comprehensive Management of Skull Base Tumors is divided into three sections consisting of: • general principles • site specific surgery • tumor specific management Filled with scientific tables and lavishly illustrated, this text is written with an emphasis on surgery, radiation and chemotherapy, and will appeal to all neurosurgeons, otolaryngologists, plastic surgeons, maxillofacial surgeons, ophthalmologists, medical and radiation oncologists, and radiologists. about the editors... EHAB Y. HANNA is Professor of Head and Neck Surgery and Neurosurgery, The University of Texas M.D. Anderson Cancer Center, and Adjunct Professor of Otolaryngology, Baylor College of Medicine, Houston, Texas, USA. Dr. Hanna is also Vice Chair for Clinical Affairs and Medical Director of the Head and Neck Center at M.D. Anderson Cancer Center. He obtained his M.D. from Ain Shams University School of Medicine, Cairo, Egypt. Dr. Hanna currently serves on the editorial boards of several publications; he is Editor-in-Chief of Head and Neck and Editor of the Head and Neck Cancers Section of Current Oncology Reports. He is an involved member of the American Head and Neck Society, the North American Skull Base Society, the American Academy of Otolaryngology-Head and Neck Surgery, the American College of Surgeons, and the National Cancer Institute. Dr. Hanna has been nationally recognized as one of “America’s Top Doctors” and “America’s Top Doctors for Cancer”. He has been a frequent invited guest speaker at national and international conferences and lectures and has contributed to numerous peer-reviewed publications in the fields of head and neck cancer and skull base tumors. FRANCO DEMONTE is Professor of Neurosurgery and Head and Neck Surgery, Deputy Chairman of the Department of Neurosurgery, and Medical Director of the Brain and Spine Center, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, USA. Dr. DeMonte is also an adjunct professor of neurosurgery at Baylor College of Medicine, Houston, Texas, USA. He received his M.D. from the University of Western Ontario, London, Canada. He is an active member of several organizations, including the Society of Neurological Surgeons, the American Association of Neurological Surgeons, and the Canadian Neurosurgical Society. He has served as president of the North American Skull Base Society and of the Houston Neurological Society. His clinical and educational activities have been recognized through his national and international presentations as well as by several teaching awards and inclusion in “Best Doctors in America” and “America’s Top Doctors”. Printed in the United States of America
Comprehensive Management of Skull Base Tumors
Otolaryngology/Neurosurgery
Comprehensive Management of Skull Base Tumors Edited by
Ehab Y. Hanna & Franco DeMonte
Hanna DK054X
•
DeMonte